Method and system for monitoring rubber coagulation reaction based on Raman spectrum and medium

Abstract

The invention relates to the field of production and processing of rubber, in particular to optimization of a monitoring technology for a rubber coagulation reaction. The embodiment of the invention discloses a method for monitoring rubber coagulation reaction based on Raman spectrum. The method comprises the following steps: collecting Raman signal light reflected back from a rubber coagulation reaction device; identifying and extracting a specific Raman characteristic peak according to the collected Raman signal light; based on the Raman characteristic peak, identifying a to-be-detected substance and a concentration value of the to-be-detected substance; and controlling the rubber coagulation reaction device in real time based on the to-be-detected substance and the concentration value of the to-be-detected substance. The invention aims to provide the method, the system and the medium for monitoring the rubber coagulation reaction based on the Raman spectrum, in-situ and on-line monitoring of key process parameters is realized, so that production control personnel can almost grasp the dynamic state of the reaction in real time, and production management is changed from a post-operation mode based on historical data to a post-operation mode based on historical data to a post-operation mode based on historical data to a post-operation mode based on historical data to a post-operation mode based on historical data. And a timely regulation and control mode based on real-time data is improved.

Description

Technical Field

[0001] This invention relates to the field of rubber production and processing, and more specifically to the optimization of monitoring technology for rubber coagulation reactions. Background Technology

[0002] Rubber production, such as the production of halogenated butyl rubber, is an important polymer synthesis process. Its essence is the chemical modification of the base polymer, butyl rubber. Butyl rubber is copolymerized from isobutylene and a small amount of isoprene, exhibiting high saturation of its molecular backbone and excellent inherent gas tightness. The purpose of the halogenation reaction is to introduce highly reactive halogen atoms, such as bromine or chlorine, into the butyl rubber molecular chain. This modification process significantly increases the vulcanization rate of the rubber and improves its compatibility and adhesion with other unsaturated rubbers. Therefore, halogenated butyl rubber is widely used as a key raw material in the manufacture of high-performance tire inner linings, and is a core material ensuring tire safety, durability, and energy efficiency.

[0003] During the aforementioned production process, rubber concentration monitoring is necessary. Rubber concentration monitoring specifically refers to the quantitative analysis of the mass fraction, or concentration, of the base butyl rubber solution dissolved in the organic solvent. This monitoring is a critical control parameter in the production process. Because the butyl rubber solution is the main raw material for the subsequent halogenation reaction, the accuracy and stability of its concentration directly determine the precision of the reactant ratio, as well as the reaction rate and uniformity.

[0004] In existing technologies, the monitoring of adhesive concentration is often an offline operation. This involves manually extracting representative samples from the operating reaction apparatus and then sending the samples to a central laboratory far from the production site, where specialized instruments such as the heating loss method or gravimetric method are used. For testing halogen content and calcium stearate content, a fluorescence spectrometer is used.

[0005] However, offline monitoring has certain technical limitations, including a significant lag. The entire process, from sampling, sample delivery, queuing, analysis to data reporting, typically takes a considerable amount of time, such as more than two hours. Production personnel cannot react immediately to fluctuations in the current moment. When the delayed detection data shows that the rubber concentration has deviated from the target value, the production unit has already been operating under suboptimal conditions for a long time. Process adjustments at this point, such as adding solvent or rubber, are delayed remedial measures, which can lead to quality fluctuations throughout the batch and even the production of defective products. Summary of the Invention

[0006] The purpose of this invention is to provide a method, system, and medium for monitoring the coagulation reaction of rubber based on Raman spectroscopy. This enables in-situ, online monitoring of key process parameters, allowing production control personnel to grasp the reaction dynamics in near real-time. This transforms production management from a post-event operation mode based on historical data to a timely control mode based on real-time data.

[0007] According to a first aspect of this disclosure, a method for monitoring the coagulation reaction of rubber based on Raman spectroscopy is provided, the method comprising: Collect the Raman signal light emitted from the rubber coagulation reaction device; Based on the collected Raman signal light, specific Raman characteristic peaks are identified and extracted; Based on the Raman characteristic peaks, the analyte and its concentration value are identified; The rubber coagulation reaction device is controlled in real time based on the substance to be tested and its concentration value.

[0008] According to a second aspect of this disclosure, a device for monitoring the coagulation reaction of rubber based on Raman spectroscopy is provided, the device comprising: The collection module is configured to collect the Raman signal light emitted from the rubber coagulation reaction device; The feature peak identification module is configured to identify and extract specific Raman feature peaks based on the collected Raman signal light; The judgment module is configured to identify the analyte and its concentration value based on the Raman characteristic peaks. The reporting module is configured to perform real-time control of the rubber coagulation reaction device based on the analyte and its concentration value.

[0009] According to a third aspect of this disclosure, an electronic device is provided, including one or more processors and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform a method provided according to a first scheme.

[0010] According to a fourth aspect of this disclosure, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided according to the first aspect.

[0011] It should be understood that the description in the Summary of the Invention section is not intended to limit the key or essential features of the embodiments of this disclosure, nor is it intended to restrict the scope of this disclosure. Other features of this disclosure will become readily apparent from the following description. Attached Figure Description

[0012] Figure 1 A schematic diagram of an example environment in which several embodiments of this specification may be implemented is shown; Figure 2 Flowcharts illustrating several embodiments of the method for monitoring rubber coagulation reaction based on Raman spectroscopy in this specification are shown. Figure 3 Specific flowcharts illustrating various embodiments of this specification for obtaining concentration data are shown; Figure 4 An example block diagram of a device for monitoring the coagulation reaction of rubber based on Raman spectroscopy, according to some embodiments of this specification, is shown; Figure 5 A schematic diagram of the structure of an electronic device according to some embodiments of this specification is shown. Detailed Implementation

[0013] The present invention will be further described in detail below with reference to the accompanying drawings.

[0014] The technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings.

[0015] The terms "first," "second," "third," etc., in the description, claims, and accompanying drawings are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such processes, methods, products, or apparatus.

[0016] The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes may be made to the function and arrangement of the described elements without departing from the scope of this specification. Various processes or components may be appropriately omitted, substituted, or added to the examples. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Furthermore, features described with respect to some examples may be combined into other examples.

[0017] As mentioned earlier, in the typical solution-based production process of halogenated butyl rubber, rubber concentration detection specifically refers to the quantitative analysis of the mass fraction, or concentration, of the base butyl rubber solution dissolved in an organic solvent, such as hexane. More specifically, it monitors the mass of butyl rubber contained per unit volume of solution in the reaction system.

[0018] This test is a critical control parameter in the production process. Because the butyl rubber solution is the main raw material for the subsequent halogenation reaction, the accuracy and stability of its concentration directly determine: the precision of the reactant ratio (the amount of halogenating agent, such as bromine, needs to be precisely calculated based on the rubber concentration); and the reaction rate and uniformity (concentration directly affects the viscosity and mass transfer efficiency of the reaction system, thus influencing the progress and effect of the halogenation reaction).

[0019] Therefore, the concentration of the adhesive is the primary prerequisite for ensuring that the chemical reaction proceeds as expected and that the final product has uniform and stable performance.

[0020] While offline monitoring can provide accurate data, its inherent mode has the following serious drawbacks, especially in modern, efficient, and continuous chemical production: First, there is a significant lag: this is the most critical flaw. From sampling, sample delivery, queuing, analysis to data reporting, the entire process typically takes several hours. This means that production personnel are constantly adjusting to the "past" production status and cannot react instantly to fluctuations in the current moment.

[0021] Second, the delay in guiding process adjustments: When delayed test data shows that the rubber concentration has deviated from the target value, the production unit has already been operating under suboptimal conditions for a long time. At this point, process adjustments, such as adding solvent or rubber, are belated remedial measures, which may lead to quality fluctuations throughout the batch, or even the production of defective products.

[0022] Third, the issue of sample representativeness: manual sampling may be biased, especially for viscous fluids such as butyl rubber solutions, which are prone to localized uneven concentrations. Single-point sampling may not accurately reflect the average concentration of the entire reaction system.

[0023] Fourth, high labor intensity and safety risks: Frequent manual sampling increases the labor intensity of operators and exposes them to potentially hazardous chemical environments.

[0024] Fifth, it cannot achieve process optimization and closed-loop control: because the data is discrete and discontinuous, it cannot capture transient and continuous changes in the production process, and therefore cannot support advanced real-time process optimization and fully automatic closed-loop control.

[0025] In conclusion, the limitations of offline testing make it a key bottleneck restricting the stability and efficiency improvement of halogenated butyl rubber production quality.

[0026] Therefore, embodiments of this disclosure propose a method for monitoring the coagulation reaction of rubber based on Raman spectroscopy. In these embodiments, the method differs from offline monitoring by employing an online monitoring mode. By directly and in-situ installing the monitoring probe on the reaction device, real-time and continuous signal acquisition of the reaction process is achieved, and an intelligent algorithm model is used to instantly convert the spectral data into key process parameters. This reduces the data response time from hours to seconds, transforming the production process from relying on lagging historical data for judgment to precise control based on real-time data.

[0027] Figure 1 A schematic diagram of an example environment 100 in which several embodiments of the present disclosure may be implemented is shown.

[0028] like Figure 1As shown, the environment 100 mainly includes a reaction device 101, a data acquisition device 102, a computing system 103, a display device 104, and a training module 105.

[0029] The reaction device 101 is the core container for the coagulation reaction, where the chemical reaction process of the rubber materials takes place. The acquisition device 102 is directly installed at the detection port of the reaction device 101, using an in-situ installation method, to acquire Raman spectral signals within the reaction system in real time. This probe is a composite optical detection module specifically designed for extreme chemical environments. Its front end is equipped with a quartz protective window and temperature control function, enabling it to withstand harsh conditions of high temperature and high pressure, achieving direct, non-contact measurement of the reaction process.

[0030] The computing system 103 is connected to the acquisition device 102 via cable or wirelessly to receive the spectral data transmitted from it. This system includes a signal processing module and a process feedback module, with a built-in embedded analysis unit. It runs a trained quantitative analysis model to perform a series of processes on the received raw spectral data, including dynamic baseline correction, spectral stitching, characteristic peak extraction, and partial least squares (PLS) regression analysis, ultimately retrieving the concentrations of key components in the reaction system in real time.

[0031] The computing system 103 can be any device with computing or processing capabilities. For example, the computing system can include, but is not limited to, mobile phones, tablets, desktop computers, servers, etc.

[0032] In one or more embodiments of this disclosure, the computing system may include a central processing unit (CPU) responsible for general computing tasks and system control, a graphics processing unit (GPU) dedicated to parallel computing and graphics rendering, a memory subsystem (including system memory and video memory), an input / output interface for data exchange, and a storage device for program and data persistence.

[0033] The training module 105 is used in the initial system setup phase. It consists of a series of standard samples with known precise concentrations, used to train and calibrate the analysis model in the computing system 103 to establish a high-precision quantitative analysis model.

[0034] The display device 104 is connected to the computing system 103 and is used to visualize the monitoring results. It displays the change curve of reactant concentration and process status in real time, and provides early warning information when the concentration deviates from the set threshold, providing operators with an intuitive monitoring interface.

[0035] The system's workflow is as follows: The acquisition device 102 continuously monitors the condensation reaction process within the reaction device 101, and the acquired spectral signals are transmitted to the computing system 103. The computing system 103 uses a pre-trained model to perform real-time analysis of the spectrum to obtain accurate concentration data. Finally, the concentration information, trend graph, and early warning signals are all presented on the display device 104, and the data can be further fed back to the control system to achieve automatic adjustment of process parameters. The display device 104 can be a computer monitor, television screen, projection screen, monitoring display screen, mobile phone display screen, etc.

[0036] It should be understood that Figure 1 The architecture of the example environment 100 shown is primarily for illustrating the core concepts of the embodiments of this disclosure, and not for limiting the scope of this disclosure. The embodiments of this disclosure can also be applied to other similar chemical process monitoring scenarios.

[0037] Figure 2 A flowchart of a method for monitoring rubber coagulation reaction based on Raman spectroscopy according to several embodiments of this specification is shown. The specific operation steps of this method 200 are as follows.

[0038] In box 201, method 200 allows for the deployment of hardware devices. This step requires the rational, stable, and safe integration of a series of hardware devices specifically designed for extreme chemical environments into the production unit, based on the production process flow, to construct a data sensing network capable of reliably acquiring raw signals. The deployment quality at this stage directly determines the long-term stability, accuracy, and safety of the entire monitoring system. The hardware devices to be deployed include lasers, optically coupled probes, and temperature control units.

[0039] The laser is a standalone module. In some embodiments, it is a near-infrared semiconductor laser, such as one with a wavelength of 785 nm, which minimizes interference from the rubber sample's own fluorescence. The laser beam generated by the laser is transmitted through a single-core or multi-core fiber to a fiber-coupled probe installed in the production area. The fiber-coupled probe has a robust industrial-grade metal housing, typically made of corrosion-resistant materials such as stainless steel or Hastelloy. The front end of the fiber-coupled probe is fitted with a protective window made of sapphire or quartz, which possesses extremely high hardness, chemical inertness, and excellent optical transparency, enabling it to withstand the erosion and corrosion of the process media.

[0040] The fiber optic coupled probe may contain a laser collimating lens, a collecting lens, and a high-efficiency filter. The laser collimating lens focuses the laser light from the fiber into parallel light, or focuses it to a preset fixed focal length of 1 cm. The collecting lens collects the scattered Raman signal light. The high-efficiency filter is a set of precision filters used to strictly block the reflected light from the excited laser, allowing only weak Raman signals to pass through, thus facilitating the acquisition of high-quality spectra.

[0041] The temperature control unit can be a module integrated inside the probe housing or mounted close to the probe. In some embodiments, it can be a Peltier effect cooler / heater, working in conjunction with a thermistor for precise temperature sensing and feedback control. For ultra-high temperature applications, a circulating water or gas cooling system may also be used, removing heat through coolant channels on the probe. The temperature control unit actively maintains a constant temperature for the optical components inside the probe. On one hand, it protects the optical components, preventing damage or performance degradation of lenses and filters due to overheating. On the other hand, it ensures measurement stability, as temperature fluctuations can cause slight thermal expansion and contraction of the optical components, altering the optical path and causing spectral signal drift. Temperature control ensures the long-term stability of the optical system, a crucial guarantee for the accuracy of subsequent quantitative analysis.

[0042] In some preferred embodiments, the placement location of the probe is selected.

[0043] Based on considerations of the halogenated butyl rubber production process, fiber optic coupled probes are installed at representative key workstations. For example, these are placed in raw material rubber solution delivery pipelines or at the outlet of the dissolving reactor to monitor the uniformity of the initial rubber concentration; and at the inlet or circulation pipeline of the halogenation reactor to monitor the material status entering the core reaction unit in real time. This ensures that the monitoring data directly reflects process parameters crucial to product quality.

[0044] In some preferred embodiments, a mixing device is added. High-viscosity polymer solutions tend to form a laminar flow state, resulting in a concentration gradient in the liquid within the pipes or reactor. Butyl rubber molecules may form microscopic or macroscopic "clumps" due to incomplete local dissolution or aggregation. This inhomogeneity means that the instantaneous signal acquired at a fixed detection point may not accurately reflect the average composition of the entire reaction system, leading to distorted detection results. In preferred embodiments, a mixing device, such as a corrugated stirrer, is integrated upstream of the probe within the process pipe or reactor. Its core function is to break the laminar flow and promote a highly turbulent flow state. The intense turbulence, through convection and shearing, disperses any potential "clumps" and thoroughly mixes fluid units of different concentrations. This ensures that the sample flowing through the probe's detection window tends to be homogeneous in both time and space, providing a representative physical basis for detection. This step improves sample consistency from the source.

[0045] In box 202, method 200 can set training materials for training the computational system. The model training phase is the cornerstone of the entire online monitoring system. This phase occurs before the system is officially put into online operation, and its core objective is to establish a mathematical prediction model that can accurately and reliably convert measured Raman spectral data into specific chemical components, such as butyl rubber concentration values, i.e., the computational system mentioned above.

[0046] In this step, specifically, a series of butyl rubber-hexane solution samples of different concentrations, covering the expected production range, are precisely prepared in a laboratory environment. For example, hundreds of samples with concentrations ranging from 15% to 30% are prepared. The true concentration value of each sample is accurately determined by analytical methods (such as gravimetric analysis) and used as the standard concentration value or true value for that sample.

[0047] Subsequently, using a Raman spectrometer consistent with the online system specifications, spectra of each standard sample of known concentration were acquired under controlled laboratory conditions. Preferably, this process can simulate certain conditions of the online detection, such as stirring, to ensure the representativeness of the spectra. Multiple spectra were acquired for each sample to assess repeatability. A series of algorithms were applied to the acquired raw spectra to identify and extract specific Raman characteristic peaks related to the characteristic vibrations of the butyl rubber molecular structure. The intensity or area of ​​these characteristic peaks will be closely related to the concentration. Intelligent software will precisely define the range of these characteristic peaks, achieving effective separation from characteristic peaks of other substances such as solvents, avoiding interference.

[0048] At this point, the training materials contain a large amount of training data, with each set of training data including the mapping relationship between substances, characteristic peaks, and concentrations. The computational system is continuously trained using this training data. Further, a model validation phase can be implemented. Reserved samples not used in model training are used to evaluate the model's predictive performance. The accuracy, stability, and generalization ability of the model are validated by calculating metrics such as the coefficient of determination (R², >0.99) between predicted and true concentrations and prediction errors.

[0049] In summary, within box 202, a large number of known "sample-answer" pairs were used to train the computational system to interpret complex Raman spectra, enabling it to not only identify the presence of substances but also accurately determine their concentrations. This validated, high-precision model serves as a prerequisite and theoretical basis for the reliable execution of all subsequent real-time monitoring and automatic control actions.

[0050] Boxes 201 and 202 above have a certain degree of step independence, and their time order is not unique. They can be performed sequentially or simultaneously.

[0051] In box 203, method 200 can acquire data in real time. The core task is to automatically and continuously capture the raw Raman scattering signal of the reaction system at a preset rhythm that matches the process dynamics, and convert it into a digital spectroscopic data stream for subsequent analysis.

[0052] Specifically, the data acquisition process is synchronized with the process conditions. For example, the system ensures that signal acquisition only begins or continues after the corrugated stirrer has been activated and the fluid has reached a stable turbulent state, thus guaranteeing the representativeness of the acquired samples. The acquisition module is set to acquire a wavelength of 100 cm⁻¹. -1 Up to 3800cm -1 The Raman shift signal is within a certain range. This range fully covers the main fingerprint vibrational regions of organic molecules (such as CC, CH, C=C, etc.), ensuring that the captured spectral information is sufficient for qualitative and quantitative analysis of butyl rubber, solvents, and other related components.

[0053] In some embodiments, the probe acquires data at a high frequency of 30 Hz, i.e., 30 raw spectra per second. This high-frequency setting serves two key purposes: first, it captures rapid dynamics, enabling the tracking of rapid mixing or reaction processes that may occur in the reaction system; second, it provides a robust statistical foundation, offering a large number of data points for subsequent mathematical processing, and effectively suppressing random noise or transient inhomogeneities that may be encountered in a single measurement through statistical averaging, thereby improving the robustness of the results.

[0054] Subsequently, the Raman light signal collected by the optical system within the probe is transmitted via optical fiber to the spectrometer in the safe zone. The spectrometer's dispersive system has a resolution ≤2 cm⁻¹. -1 The composite light is dispersed into a spectrum, which is then converted into a high-precision digital signal by a charge-coupled device (CCD) detector, namely a series of digital spectra containing the relationship between intensity and wavenumber.

[0055] Box 204 is one of the core steps of this scheme. In box 204, the computing system uses a pre-trained model to perform real-time analysis of the spectrum to obtain accurate concentration data.

[0056] Figure 3 The specific implementation of block 204 in method 200 is shown.

[0057] In box 2041, the spectrum is preprocessed to eliminate noise. Specifically, a dynamic baseline correction algorithm based on wavelet transform is employed. This method differs from traditional static or polynomial fitting baseline subtraction. Wavelet transform can analyze signals simultaneously in the frequency and time domains, intelligently identifying and separating the broad, gradually changing fluorescence background (low-frequency signal) from the sharp Raman characteristic peaks (high-frequency signal). By filtering out the low-frequency background, the algorithm can accurately flatten the spectral baseline, clearly revealing previously submerged Raman peaks, laying the foundation for subsequent quantitative analysis. This eliminates strong fluorescence background and random noise, a means of obtaining high-quality Raman spectra from rubber-like samples.

[0058] Rubber, its raw materials, and many additives produce strong fluorescence under laser irradiation. This fluorescence signal intensity is stronger than the Raman signal. This causes the measured spectral baseline to drift high, potentially masking important characteristic peak information and making analysis impossible.

[0059] In some preferred embodiments, a combination of near-infrared and wavelet transform is employed. At the hardware level, near-infrared lasers have lower energy than common visible light lasers, significantly reducing the likelihood of intense fluorescence from rubber molecules and mitigating interference at its source. Wavelet transform, on the other hand, filters out low-frequency background noise at the software level. The combination of these two methods eliminates strong fluorescence background and random noise.

[0060] Box 2042 requires stitching together the filtered spectrum.

[0061] This step is to integrate data from different light sources or detector channels to expand the effective spectral range or improve the signal-to-noise ratio, forming a complete spectral dataset.

[0062] Specifically, when the system uses multiple lasers or segmented detection, the spectra of different segments need to be normalized to eliminate systematic errors caused by differences in light source intensity or detector response. Subsequently, smooth stitching is performed in the overlapping areas to generate a spectrum starting from 100 cm⁻¹. -1 up to 3800 cm -1 A continuous, seamless wide spectrum ensures the integrity of all chemical information.

[0063] Box 2043 is used for characteristic peak identification.

[0064] This step is to identify specific spectral characteristics that are directly related to the concentration of the target component, such as butyl rubber.

[0065] Based on the training described above, the computing system automatically identifies and locates characteristic Raman peaks representing the vibrations of key chemical bonds in butyl rubber molecules. Subsequently, the intensity, area, or peak shape parameters of these characteristic peaks are quantified. The software can effectively distinguish the target peak from peaks of other components (such as solvent hexane and water) and possible impurity peaks, avoiding misjudgments. The identification results not only identify the components of the target substance but also its concentration data.

[0066] In some preferred embodiments, concentration inversion is included.

[0067] In industrial online monitoring, ambient temperature fluctuations are a significant source of interference affecting measurement accuracy. Temperature changes can lead to spectral peak shifts, peak shape variations, and changes in background intensity. These effects can alter the spectral shape of the same sample measured at different temperatures. Theoretically, model identification will introduce certain errors when operating temperature fluctuates.

[0068] In this step, a partial least squares model with temperature compensation was used.

[0069] Specifically, first, the pre-trained PLS model parameters, which include temperature compensation information, are loaded. These parameters include the core mathematical framework such as the number of latent variables, loading vectors, and regression coefficient vectors. Simultaneously, the spectral data from the previous stage, which has undergone preprocessing, dynamic baseline correction, and spectral stitching, is sent to the model waiting queue.

[0070] Subsequently, the system projects this spectral vector onto the feature space obtained during model training. Specifically, it substitutes the spectral data into a new coordinate system defined by the loading matrix obtained during model training, and calculates the scores of the new spectrum on each latent variable.

[0071] Next, the concentration identification and temperature compensation are calculated simultaneously.

[0072] The concentration calculation involves multiplying the latent variable scores from the new spectrum by the regression coefficient vector connecting the latent variable scores and the concentration values, obtained during model training, to arrive at a preliminary concentration identification value without temperature correction. Temperature interference estimation occurs at this point. Since the model is trained on a dataset containing temperature variations, some latent variables inherently carry strong information about temperature changes. The system identifies these temperature-related latent variables and their scores, and estimates the amount of deviation in concentration prediction caused by the current temperature conditions based on the "temperature-spectral response relationship" learned within the model.

[0073] The final concentration value is generated by subtracting the temperature interference deviation from the initial concentration identification value. This corrected concentration value is the final result derived from the model inversion.

[0074] It should be noted that the PLS partial least squares regression model is used here. On the one hand, it can improve the accuracy of concentration measurement. On the other hand, it is used in conjunction with the corrugated stirrer mentioned above to solve the problem of uneven rubber solution in the solution production process of halogenated butyl rubber.

[0075] This is because the large amount of instantaneous spectral data collected is input into a pre-established partial least squares regression model for processing. Chemometric models such as PLS do not derive reliable conclusions from a single measurement, but rather perform mathematical averaging and statistical learning based on a large amount of spectral data over a period of time. This results in the model's final output concentration value not being an instantaneous value, but an optimized statistical value based on a large number of data points. It effectively filters out random fluctuations caused by microscopic inhomogeneities, reflecting the statistically average concentration of components over that time period, thus providing extremely stable and reliable results.

[0076] Finally, proceed to box 205, where method 200 can perform control operations based on the concentration value of the target reference obtained above.

[0077] When the system determines, through real-time analysis, that the concentration of the target component deviates from the preset process index range, such as ±2%, it will automatically trigger a series of programmed response actions, which aim to correct the deviation in a timely manner, alert operators, and retain key data.

[0078] In some embodiments, this control operation involves adjusting the reactor feed flow rate via a PID controller. Specifically, the system uses the real-time concentration deviation value—the difference between the setpoint and the measured value—as an input signal to the control algorithm of the proportional-integral-derivative (PID) controller. The controller automatically and accurately calculates adjustment commands based on the magnitude, duration, and trend of the deviation, and sends them to the control valve (such as a regulating valve) on the reactor feed line, thereby changing the feed flow rate of the relevant material.

[0079] In some embodiments, this control operation involves updating the HMI (Human Machine Interface) warning information. Specifically, while automatically adjusting the process, the system sends instructions to the HMI to update the displayed content. On the corresponding flowchart screen, the color of the abnormal point will change, such as turning yellow or red, and a flashing icon may appear. Simultaneously, a warning window pops up, clearly listing the abnormal parameter, current value, setpoint, and degree of deviation. The warning event is recorded in the system's alarm log.

[0080] In some embodiments, the control operation may also include storing abnormal spectral data for quality traceability. The system automatically packages the raw spectral data and related process parameters collected within a period before and after the triggering of the abnormal event, adds a timestamp tag, and stores it in a specific database or historical data file, which enables in-depth analysis.

[0081] In some embodiments, a genetic algorithm is used to calculate the optimal installation method for the probe. Specifically, the probe's position, i.e., its coordinates in three-dimensional space, and the probe's angles, such as pitch and yaw angles, are used as decision variables for the genetic algorithm.

[0082] The fitness formula is: Fitness = w1·f_SI + w2·f_CR + w3·f_HD - w4·f_DR.

[0083] Among them, w1 to w4 are weight functions, which are user-defined.

[0084] There are four constraints here: A high signal strength index f_SI means maximizing signal strength and stability, ensuring that the detection system obtains high-quality spectral data; A high concentration representativeness index (f_CR) ensures that the concentration value measured by the probe can accurately reflect the true state of the entire reaction system. A high fluid dynamics index f_HD indicates the selection of a location with appropriate flow velocity, stable turbulence, and avoidance of bubble interference. The dead zone risk index f_DR is low, thus avoiding the installation of the probe in the flow dead zone or material stagnation area.

[0085] The model selected is a fluid dynamics CFD model. A genetic algorithm is used for traversal calculations. The solution with the highest fitness result is the optimal solution. The probe position and angle under the optimal solution are the optimal installation position and angle.

[0086] It should be noted that the substance to be tested can be different chemical substances depending on the project and environment, including but not limited to halogens and calcium stearate. It is necessary to obtain the halogen content and calcium stearate content.

[0087] Figure 3 A schematic diagram of the structure of a monitoring device 300 for monitoring the coagulation reaction of rubber based on Raman spectroscopy, according to some embodiments of this disclosure, is shown. The various embodiments in this specification are described in a progressive manner; similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, for the device embodiments, since they are basically similar to the method embodiments, the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments. Figure 3 As shown, the device 300 includes a collection module 301 configured to collect Raman signal light emitted from the rubber coagulation reaction device; a characteristic peak identification module 302 configured to identify and extract specific Raman characteristic peaks based on the collected Raman signal light; a judgment module 303 configured to identify the analyte and its concentration value based on the Raman characteristic peaks; and a reporting module 304 configured to perform real-time control of the rubber coagulation reaction device based on the analyte and its concentration value.

[0088] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this specification are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., Digital Versatile Discs (DVDs)), or semiconductor media (e.g., Solid State Disks (SSDs)).

[0089] Figure 4 A block diagram of an electronic device 400 that can implement various embodiments of the present disclosure is shown. For example... Figure 4 As shown, the electronic device 400 includes a processor 401, which can perform various appropriate actions and processes based on computer program instructions loaded into random access memory (RAM) 403 according to computer program instructions stored in read-only memory (ROM) 402. The RAM 403 may also store various programs and data required for the operation of the device 400. The processor 401, ROM 402, and RAM 403 are interconnected via a bus 404. An input / output (I / O) interface 405 is also connected to the bus 404.

[0090] The various processes and procedures described above, such as method 300, can be executed by processor 401. For example, in some embodiments, method 100 may be implemented as a software program tangibly contained in a machine-readable medium. In some embodiments, part or all of the software program may be loaded and / or installed on device 400 via ROM 402. When the software program is loaded into RAM 403 and executed by processor 401, one or more actions of method 200 described above may be performed.

[0091] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0092] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Furthermore, although operations are depicted in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the foregoing discussion, these should not be construed as limiting the scope of this disclosure. Certain features described in the context of individual embodiments may also be implemented in combination in a single implementation. Conversely, various features described in the context of a single implementation may also be implemented individually or in any suitable sub-combination in multiple implementations.

[0093] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method for monitoring the coagulation reaction of rubber based on Raman spectroscopy, characterized in that, The method includes: Collect the Raman signal light emitted from the rubber coagulation reaction device; Based on the collected Raman signal light, specific Raman characteristic peaks are identified and extracted; Based on the Raman characteristic peaks, the analyte and its concentration value are identified; The rubber coagulation reaction device is controlled in real time based on the substance to be tested and its concentration value.

2. The method for monitoring rubber coagulation reaction based on Raman spectroscopy according to claim 1, characterized in that: During the process of collecting the Raman signal light emitted from the rubber coagulation reaction device, The collection device includes a laser and a light-coupled probe, wherein the laser is a near-infrared laser.

3. The method for monitoring rubber coagulation reaction based on Raman spectroscopy according to claim 2, characterized in that: In the process of identifying and extracting specific Raman characteristic peaks based on the collected Raman signal light, The wavelet transform dynamic baseline correction algorithm is used to eliminate noise in the spectrum of Raman signal light and flatten the spectral baseline.

4. The method for monitoring rubber coagulation reaction based on Raman spectroscopy according to claim 1, characterized in that, The method further includes: in the process of identifying and extracting specific Raman characteristic peaks based on the collected Raman signal light, The spectra from different light sources or detector channels are normalized and then stitched together to form a complete spectral dataset.

5. The method for monitoring rubber coagulation reaction based on Raman spectroscopy according to claim 1, characterized in that, The method further includes: before collecting the Raman signal light emitted from the rubber coagulation reaction device, A mixing device is used to stir the fluid to be tested, thereby creating turbulence in the adhesive solution; In the process of identifying the analyte and its concentration value based on the Raman characteristic peak, a PLS partial least squares regression model containing temperature compensation information is used to obtain the final concentration value after temperature correction.

6. The method for monitoring rubber coagulation reaction based on Raman spectroscopy according to claim 1, characterized in that, Before the collection of Raman signal light emitted from the rubber coagulation reaction device, Install the optically coupled probe in the raw material adhesive delivery pipeline and / or the outlet of the dissolving vessel and / or the inlet and / or the circulation pipeline of the halogenation reactor.

7. The method for monitoring rubber coagulation reaction based on Raman spectroscopy according to claim 1, characterized in that, Before collecting the Raman signal light emitted from the rubber coagulation reaction device, the optimal arrangement of the light-coupled probes is calculated based on the genetic algorithm (GA). The decision variables are set as the angle value and the coordinate value of the probe in three-dimensional space. The constraints are set as high signal intensity index, high concentration representativeness index, high fluid dynamics index, and low dead zone risk index.

8. A gesture 3D reconstruction device based on binocular vision, characterized in that, The device includes: The collection module is configured to collect the Raman signal light emitted from the rubber coagulation reaction device; The feature peak identification module is configured to identify and extract specific Raman feature peaks based on the collected Raman signal light; The judgment module is configured to identify the analyte and its concentration value based on the Raman characteristic peaks. The reporting module is configured to perform real-time control of the rubber coagulation reaction device based on the analyte and its concentration value.

9. An electronic device, comprising: One or more processors, and A memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method for monitoring rubber coagulation reaction based on Raman spectroscopy as described in any one of claims 1-7.

10. A computer program product, comprising a computer program that, when executed by a processor, implements the method for monitoring the coagulation reaction of rubber based on Raman spectroscopy according to any one of claims 1-7.

Images