A solution concentration online monitoring method and system based on refractive index

Abstract

The application relates to the technical field of industrial automation control, and provides a solution concentration online monitoring method and system based on refractive index, which comprises the following steps: obtaining a refractive index signal of a refractive index sensor, a physical disturbance signal of a physical disturbance sensor, and a temperature distribution signal of a temperature probe; identifying a physical disturbance component according to the time corresponding relationship between the physical disturbance signal and the refractive index signal; identifying a temperature characteristic region based on the temperature distribution signal, and generating a pre-judgment moment according to the temperature characteristic region; monitoring the refractive index signal based on the pre-judgment moment, and recording the current change as a refractive index change when the refractive index signal changes correspondingly; and temperature compensating the refractive index change by using the temperature distribution signal to obtain information representing the solution concentration. The application has the advantages of fast and accurate capturing and feedback of the solution concentration change, and significantly improves the reliability and response timeliness of online monitoring.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of industrial automation control, and in particular to a solution concentration online monitoring method and system based on refractive index. BACKGROUND

[0002] In the chemical industry production link, accurately and instantly grasping the concentration of the material components in the reaction kettle is the basis for ensuring the uniformity of product quality and optimizing the production process. When deploying and running such online monitoring systems aimed at improving response timeliness in practice, a series of challenges originating from real industrial environments will be faced.

[0003] Firstly, even if a refractive index sensing element with extremely short response time is selected, when dealing with some materials with special physical properties, such as polymer solutions with large viscosity values or slurry containing a certain amount of incompletely separated suspended solids, the flow behavior of these materials in the pipeline itself will limit the update rate of the sample in the sensor sensing area, resulting in time delay in physical transportation. Secondly, the measured value of the refractive index, a physical quantity, shows significant dependence on the temperature change of the measured medium. When the temperature change amplitude of the solution in the main flow is large and the change rate is fast, the response capability and control precision of the temperature compensation system itself may not fully match the dynamic process of the actual temperature change, introducing measurement errors. Furthermore, in the continuous flow of the solution, some small solid particles or micro-bubbles may be entrained, and these out-of-phase particles or bubbles will cause scattering, non-uniform absorption or irregular refraction of the incident light when they flow with the solution to the optical sensing interface of the refractive index sensor, resulting in short-term, abnormal amplitude spike-like fluctuations or background noise on the refractive index measurement signal.

[0004] In view of the above problems, the prior art needs to be improved. SUMMARY

[0005] The purpose of the present application is to provide a solution concentration online monitoring method and system based on refractive index, which can effectively solve the interference problems caused by physical transportation delay, temperature change and physical disturbance on the refractive index measurement signal, thereby realizing rapid and accurate capture and feedback of the solution concentration change, and significantly improving the reliability and response timeliness of online monitoring.

[0006] The present application provides a solution concentration online monitoring method based on refractive index, and the technical solution is as follows:

[0007] A solution concentration online monitoring method based on refractive index, comprising:

[0008] acquire a refractive index signal of a refractive index sensor arranged along a continuous flow path in the reaction kettle, a physical disturbance signal of a physical disturbance sensor, and a temperature distribution signal of a temperature probe arranged upstream of the refractive index sensor;

[0009] According to the time correspondence relationship between the physical disturbance signal and the refractive index signal, the physical interference component in the refractive index signal is identified;

[0010] Based on the temperature distribution signal, a temperature characteristic region propagating along the continuous flow path in the reaction kettle is identified, and a pre-judgment time when the material state change reaches the refractive index sensor is generated according to the temperature characteristic region;

[0011] Based on the pre-judgment time, the refractive index signal is monitored, and when the refractive index signal changes corresponding to the pre-judgment time, the current change is recorded as a refractive index change caused by solution concentration change;

[0012] The temperature distribution signal is used to temperature compensate the refractive index change to obtain information representing the solution concentration.

[0013] Through the above scheme, the influence of physical interference and temperature change on the refractive index signal can be effectively identified and removed, so as to realize rapid and accurate capture of solution concentration change, and improve the reliability and response timeliness of online monitoring.

[0014] Optionally, the application also proposes that the step of identifying a temperature characteristic region propagating along the continuous flow path in the reaction kettle based on the temperature distribution signal, and generating a pre-judgment time when the material state change reaches the refractive index sensor according to the temperature characteristic region comprises:

[0015] Based on the temperature change of the temperature characteristic region, a trend signal and a detail signal are separated;

[0016] Based on a preset change criterion, the trend signal is verified to obtain a trend verification result;

[0017] The detail signal is subjected to frequency domain analysis to determine a characteristic frequency of the detail signal;

[0018] It is judged whether the characteristic frequency corresponds to the working frequency of the delivery equipment of the driving fluid to obtain a frequency correspondence result;

[0019] When the trend verification result is passed and the frequency correspondence result is corresponding, the temperature characteristic region is determined as valid, and the pre-judgment time is generated according to the propagation characteristics of the temperature characteristic region.

[0020] Through the above scheme, the accuracy of the pre-judgment time can be ensured through the verification of the temperature characteristic region, and the precision of the concentration monitoring is further improved.

[0021] Optionally, the application also proposes that, based on the temperature change of the temperature characteristic region, the step of separating the trend signal and the detail signal comprises:

[0022] Obtaining a working frequency sequence of the conveying device corresponding in time to the temperature change of the temperature characteristic region;

[0023] Based on the working frequency sequence, determining a reference frequency;

[0024] According to the reference frequency, performing time-varying processing on the temperature change to separate the trend signal and the detail signal.

[0025] Through the above scheme, the trend signal and the detail signal in the temperature change can be more accurately separated, providing a more reliable data basis for subsequent verification and analysis.

[0026] Optionally, the application also proposes that, according to the propagation characteristics of the temperature characteristic region, the step of generating a pre-judgment time comprises:

[0027] Identifying and determining, based on the first material state change event, a first time when the temperature characteristic region reaches the refractive index sensor, and determining a second time when the solution concentration change detected by the refractive index sensor occurs;

[0028] According to the first time and the second time, determining a time deviation;

[0029] Identifying and, based on the second material state change event, generating an initial pre-judgment time according to the propagation characteristics of the temperature characteristic region corresponding to the second material state change event;

[0030] Based on the time deviation, correcting the initial pre-judgment time to generate a pre-judgment time when the material state change reaches the refractive index sensor.

[0031] Through the above scheme, the pre-judgment time can be corrected through historical events, improving the accuracy and adaptability of the pre-judgment and reducing errors caused by material conveying delays.

[0032] Optionally, the application also proposes that, according to the propagation characteristics of the temperature characteristic region, the step of generating a pre-judgment time comprises:

[0033] Obtaining historical material state change events; each historical material state change event includes a historical time deviation and a historical process parameter;

[0034] Based on the historical material state change events, establishing a correspondence between the process parameters and the time deviations;

[0035] For the second material state change event, obtaining a current process parameter corresponding to the second material state change event;

[0036] According to the correspondence between the process parameter and the time deviation and the current process parameter, a verified time deviation corresponding to the second material state change event is determined;

[0037] The initial prediction time is corrected based on the verified time deviation, and a prediction time of the material state change reaching the refractive index sensor is generated.

[0038] Through the above scheme, a more precise time deviation correction model can be established based on historical process parameters, further improving the correction accuracy and universality of the prediction time.

[0039] Optionally, the present application also proposes that the step of establishing the correspondence between the process parameter and the time deviation based on the historical material state change event comprises:

[0040] Based on the historical material state change event, the numerical range of the historical process parameter in the historical material state change event is divided into several process parameter intervals;

[0041] For each process parameter interval, the historical material state change event corresponding to the historical process parameter falling into the process parameter interval is obtained from the historical material state change event, which is recorded as a historical control event;

[0042] Based on the historical time deviation of the historical control event, a time deviation representative value is calculated;

[0043] The mapping between the process parameter interval and the time deviation representative value is established as the correspondence.

[0044] Through the above scheme, a more robust correspondence between the process parameter and the time deviation can be established by dividing the process parameter interval and calculating the time deviation representative value, improving the accuracy of the correction model.

[0045] Optionally, the present application also proposes that the step of calculating the time deviation representative value based on the historical time deviation of the historical control event comprises:

[0046] The data distribution characteristics of the historical time deviation of the historical control event are determined;

[0047] According to the data distribution characteristics, a data concentration area is determined;

[0048] Based on the historical time deviation in the data concentration area, a time deviation representative value is calculated.

[0049] Through the above scheme, the data concentration area can be more accurately determined by analyzing the data distribution characteristics of the historical time deviation, so that a more representative time deviation representative value is calculated.

[0050] Optionally, the application further proposes that the step of determining the data distribution feature of the historical time deviation of the historical control event comprises:

[0051] The statistical analysis is performed on the historical time deviation of the historical control event, and statistical parameters of the data distribution are obtained as the data distribution feature.

[0052] Through the above scheme, the data distribution feature can be obtained through statistical analysis, which provides a quantitative basis for determining the concentrated region of the data, and improves the scientificity of the calculation of the time deviation representative value.

[0053] Optionally, the application further proposes that the step of determining the data distribution feature of the historical time deviation of the historical control event comprises:

[0054] Based on the data distribution feature, the density distribution of the historical time deviation is determined;

[0055] According to the density distribution, the peak value region is determined;

[0056] The peak value region is taken as the concentrated region of the data.

[0057] Through the above scheme, the peak value region can be determined through the density distribution, and the concentration trend of the time deviation can be more accurately identified, and the calculation of the time deviation representative value is further optimized.

[0058] A solution concentration online monitoring system based on refractive index is used to perform online monitoring of solution concentration based on refractive index, comprising:

[0059] A signal acquisition module is configured to acquire a refractive index signal of a refractive index sensor arranged along a continuous flow path in a reaction kettle, a physical disturbance signal of a physical disturbance sensor, and a temperature distribution signal of a temperature probe arranged upstream of the refractive index sensor;

[0060] An interference identification module is configured to identify a physical interference component in the refractive index signal according to the time corresponding relationship between the physical disturbance signal and the refractive index signal;

[0061] A pre-judgment time generation module is configured to identify a temperature characteristic region propagating along the continuous flow path in the reaction kettle based on the temperature distribution signal, and generate a pre-judgment time of the material state change reaching the refractive index sensor according to the temperature characteristic region;

[0062] A change recording module is configured to monitor the refractive index signal based on the pre-judgment time, and record the current change as a refractive index change caused by the change of the solution concentration when the refractive index signal changes corresponding to the pre-judgment time;

[0063] A concentration monitoring module is configured to perform temperature compensation on the refractive index change using the temperature distribution signal to obtain information representing the solution concentration.

[0064] The system for realizing the monitoring method is modularized, which facilitates actual deployment and operation, and improves the integration and automation level of the system.

[0065] As can be seen, the online solution concentration monitoring method and system based on refractive index provided by the application effectively identify and eliminate interference by introducing a physical disturbance signal, a temperature distribution signal and a pre-judgment timing mechanism, accurately monitor real concentration changes, effectively solve the interference problem caused by physical delivery delay, temperature change and physical disturbance on the refractive index measurement signal in the prior art, thereby realizing rapid and accurate capture and feedback of solution concentration changes, and significantly improving the reliability and response timeliness of online monitoring. BRIEF DESCRIPTION OF DRAWINGS

[0066] Figure 1 A method flowchart of an online solution concentration monitoring method based on refractive index in one embodiment of the application;

[0067] Figure 2 One of the method flowcharts of an online solution concentration monitoring method based on refractive index in another embodiment of the application;

[0068] Figure 3 The second of the method flowcharts of an online solution concentration monitoring method based on refractive index in another embodiment of the application;

[0069] Figure 4 The third of the method flowcharts of an online solution concentration monitoring method based on refractive index in another embodiment of the application;

[0070] Figure 5 The fourth of the method flowcharts of an online solution concentration monitoring method based on refractive index in another embodiment of the application;

[0071] Figure 6 The fifth of the method flowcharts of an online solution concentration monitoring method based on refractive index in another embodiment of the application;

[0072] Figure 7 The sixth of the method flowcharts of an online solution concentration monitoring method based on refractive index in another embodiment of the application;

[0073] Figure 8 The seventh of the method flowcharts of an online solution concentration monitoring method based on refractive index in another embodiment of the application;

[0074] Figure 9 A system block diagram of an online solution concentration monitoring system based on refractive index in another embodiment of the application;

[0075] BRIEF DESCRIPTION OF DRAWINGS

[0076] 1. A solution concentration online monitoring system based on refractive index; 11, signal acquisition module; 12, interference identification module; 13, pre-judgment time generation module; 14, change record module; 15, concentration monitoring module. DETAILED DESCRIPTION

[0077] The technical solutions in the application will be described clearly and completely in the following combined with the drawings in the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. The components of the application described and shown in the drawings can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the application provided in the drawings is not intended to limit the scope of the claimed application, but only represents selected embodiments of the application. Based on the embodiments of the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.

[0078] It should be noted that: similar numbers and letters represent similar items in the following drawings, so once an item is defined in one drawing, it does not need to be further defined and explained in subsequent drawings. At the same time, in the description of the application, the terms "first", "second" and the like are only used to distinguish the description, and cannot be understood as indicating or implying relative importance.

[0079] The traditional existing solution concentration online monitoring method based on refractive index, when applied to the continuous flow path outside the reaction kettle, when the pipe diameter cannot be further reduced due to process needs and the solution viscosity is high, resulting in inherent delay in sample delivery, at the same time, the heat effect of the reaction kettle itself or the change of the external environment causes the temperature of the effluent material to fluctuate greatly and rapidly, and there are still micro-bubbles or solid particles in the solution which cause physical disturbance to optical measurement. When the three adverse conditions coexist, it is difficult to effectively distinguish the real solution concentration change trend from the mixed refractive index raw readings, and it is difficult to weaken or eliminate the false signal components caused by the above-mentioned delivery delay, temperature fluctuation and physical disturbance, thereby affecting the realization of rapid capture and feedback of the actual concentration change of the solution in the main flow path under the premise of ensuring the reliability of the measurement results.

[0080] To this end, the application provides a solution concentration online monitoring method based on refractive index. A solution concentration online monitoring method based on refractive index, combined with Figure 1 As shown in the drawings, comprising:

[0081] S1, obtaining the refractive index signal of the refractive index sensor arranged along the continuous flow path in the reaction kettle, the physical disturbance signal of the physical disturbance sensor, and the temperature distribution signal of the temperature probe arranged upstream of the refractive index sensor;

[0082] S2, identifying the physical disturbance component in the refractive index signal according to the time correspondence between the physical disturbance signal and the refractive index signal;

[0083] S3, identifying the temperature characteristic region propagating along the continuous flow path in the reaction kettle according to the temperature distribution signal, and generating the pre-judgment moment of the material state change reaching the refractive index sensor according to the temperature characteristic region;

[0084] S4, monitoring the refractive index signal based on the pre-judgment moment, and recording the current change as the refractive index change caused by the solution concentration change when the refractive index signal changes corresponding to the pre-judgment moment;

[0085] S5, temperature compensation of the refractive index change by using the temperature distribution signal to obtain information representing the solution concentration.

[0086] The physical disturbance sensor refers to a device for detecting physical phenomena caused by non-concentration changes in the flow path, which can be implemented by ultrasonic sensors, conductivity sensors or visual sensors, for example by detecting bubbles, solid particles or flow rate abnormalities in the fluid, which is mainly to obtain auxiliary information related to physical interference in the refractive index signal. The temperature probe refers to a device for measuring temperature changes in the flow path, which can be implemented by thermocouple arrays, thermistor arrays or distributed optical fiber temperature sensors, for example temperature measurement at multiple points or regions along the flow path, which is mainly to obtain temperature distribution information of the material propagating in the flow path, in order to identify temperature characteristic regions and perform temperature compensation. The physical interference component refers to the signal deviation in the refractive index signal caused by non-solution concentration changes, such as physical disturbances caused by bubbles, suspended particles or flow rate fluctuations, which can be identified by signal filtering, pattern recognition or time-synchronized signal comparison techniques, which is mainly to separate the true concentration change information from the original refractive index signal. The temperature characteristic region refers to the material region propagating along the continuous flow path in the reactor, which has a specific temperature change pattern or gradient, which can be identified by data analysis of temperature sensor arrays, thermal imaging technology or fluid dynamics models, for example by analyzing temperature peaks, temperature gradients or temperature fluctuation patterns in the temperature distribution signal, which is mainly to characterize the propagation characteristics of material state changes in the flow path. The pre-judgment time refers to the time point when the material state change reaches the refractive index sensor according to the propagation characteristics of the temperature characteristic region, which can be implemented by time deviation correction based on historical data, fluid transport model calculation or machine learning prediction model, for example by analyzing the propagation time of the temperature characteristic region from the upstream temperature probe to the refractive index sensor, which is mainly to predict the time when the true concentration change signal reaches the refractive index sensor in advance, so as to realize the rapid response to the concentration change. Temperature compensation refers to correcting the refractive index change according to the temperature distribution signal to eliminate the influence of temperature on the refractive index measurement value, which can be implemented by lookup table method, mathematical model correction or neural network model, for example by establishing the functional relationship between refractive index and temperature, concentration for correction, which is mainly to ensure that the refractive index measurement value accurately reflects the solution concentration and is not disturbed by temperature fluctuations.

[0087] In some preferred embodiments, the application is implemented as follows to further illustrate the working principle described above. In terms of signal acquisition, the refractive index sensor can adopt a prism or critical angle refractometer, the physical disturbance sensor can adopt an ultrasonic flow meter or micro-bubble detector installed on the flow path, and the temperature probe can adopt a plurality of thermocouple sensors uniformly distributed along the flow path. Specifically, in identifying the physical disturbance component of the refractive index signal, a cross-correlation analysis method can be used to time-align and compare the physical disturbance signal and the refractive index signal, so as to identify and filter out the instantaneous signal spikes caused by bubbles or solid particles. As a specific implementation, in identifying the temperature characteristic region propagating along the continuous flow path in the reaction kettle, a time series analysis can be performed on the temperature distribution signal collected by the temperature probe, for example, a sliding average or wavelet transform is used to identify the starting point and propagation speed of the temperature fluctuation. Subsequently, according to the propagation characteristics of these temperature characteristic regions, for example, by calculating the time required for the temperature wave peak to propagate from the upstream probe to the refractive index sensor, the pre-judgment time of the material state change reaching the refractive index sensor is generated. For example, when the pre-judgment time comes, the system focuses on the change trend and amplitude of the refractive index signal. If the refractive index signal appears a change near the pre-judgment time consistent with the expected concentration change direction, it is confirmed as a true refractive index change caused by solution concentration change. Finally, in temperature compensation of the refractive index change using the temperature distribution signal, a three-dimensional calibration table of refractive index and temperature, concentration can be established in advance, or a polynomial regression model is used. When the refractive index change value caused by the concentration change is obtained, combined with the real-time temperature at the refractive index sensor, the temperature corrected information accurately representing the solution concentration can be obtained by table lookup or model calculation.

[0088] Optionally, in combination with Figure 2 As shown in FIG. 3, S3 identifies the temperature characteristic region propagating along the continuous flow path in the reaction kettle based on the temperature distribution signal, and generates the pre-judgment time of the material state change reaching the refractive index sensor according to the temperature characteristic region. The step includes:

[0089] S31 separates the trend signal and the detail signal based on the temperature change of the temperature characteristic region;

[0090] S32 verifies the trend signal based on the preset change criterion to obtain a trend verification result;

[0091] S33 performs frequency domain analysis on the detail signal to determine a characteristic frequency of the detail signal;

[0092] S34 determines whether the characteristic frequency corresponds to the working frequency of the delivery device of the driving fluid to obtain a frequency correspondence result;

[0093] S35, when the trend verification result is passed and the frequency corresponding result is corresponding, the temperature characteristic region is determined as valid, and a pre-judgment time is generated according to the propagation characteristics of the temperature characteristic region.

[0094] Wherein, the trend signal refers to the part in the temperature change data reflecting long-term and slow change, which usually represents the real temperature response of material state change; the detail signal refers to the part in the temperature change data reflecting short-term and rapid fluctuation, which usually contains noise, interference or high-frequency transient information, which can be realized by wavelet decomposition, empirical mode decomposition or high / low pass filtering technology. The preset change criterion refers to a series of rules or conditions for judging whether the trend signal conforms to the expected mode, which can be set based on historical data, process requirements or physical model, for example, the slope, amplitude or duration of the trend signal must be within a certain range, the purpose of which is to preliminarily screen out temperature characteristic regions that do not conform to the expectation. The frequency domain analysis refers to the method of converting the signal from the time domain to the frequency domain for analysis, through which the strength and distribution of different frequency components in the signal can be revealed, for example, fast Fourier transform (FFT) or power spectral density analysis technology can be used. The characteristic frequency refers to the frequency component in the detail signal with the most concentrated energy or the largest amplitude in the frequency domain analysis, which is usually associated with the periodic interference source causing signal fluctuation. The working frequency of the fluid conveying device refers to the periodic vibration or pulsation frequency generated by the mechanical equipment (such as pump, stirrer) in the normal operation state for pushing or stirring the fluid in the continuous flow path of the reactor. The trend verification result is passed, which means that the trend signal is identified as a temperature trend caused by the expected and real material state change after being judged by the preset change criterion. The frequency corresponding result is corresponding, which means that there is a matching relationship between the characteristic frequency of the detail signal and the working frequency of the fluid conveying device, indicating that the detail signal may be affected by the periodic interference of the equipment operation. The temperature characteristic region is determined as valid, which means that the temperature change region is confirmed to be a real reflection of material state change and not significantly affected by equipment interference after being judged by the trend verification and frequency correspondence. The propagation characteristics of the temperature characteristic region refer to the time delay, attenuation or diffusion of the temperature characteristic region in the continuous flow path of the reactor from the temperature probe position to the refractive index sensor position, which can be determined based on factors such as flow rate, pipeline length and material properties, and the purpose is to provide a basis for the generation of the pre-judgment time.

[0095] In some preferred embodiments, the application is implemented as follows: when the system obtains temperature variation data of a temperature characteristic region propagating along a continuous flow path in the reactor from the temperature probe, the temperature variation data can be first subjected to multi-scale decomposition using wavelet transform, and the low-frequency components are extracted as a trend signal and the high-frequency components are extracted as a detail signal. For example, a Daubechies wavelet basis function can be selected and three-layer decomposition can be performed to effectively separate components of different frequencies. Subsequently, the trend signal obtained by separation can be verified based on a preset variation criterion. The criterion can be set as follows: the rising or falling slope of the trend signal must be between 0.5 degrees Celsius per second and 5 degrees Celsius per second, and its duration should be greater than 5 seconds. If the slope or duration of the trend signal does not meet these conditions, the trend verification result is fail. At the same time, the detail signal is subjected to frequency domain analysis, and fast Fourier transform (FFT) can be used to obtain its power spectral density. By analyzing the power spectrum, the characteristic frequency of the detail signal, i.e., the frequency point with the most concentrated energy in the power spectrum, can be determined. Then, it is judged whether the characteristic frequency corresponds to the operating frequency of the conveying device driving the fluid. For example, if the circulating pump of the reactor operates at a speed of 1500 revolutions per minute, its operating frequency can be 25 hertz. If the characteristic frequency of the detail signal is in the range of 24.5 hertz to 25.5 hertz, the frequency correspondence result is determined as corresponding. Finally, when the trend verification result is pass and the frequency correspondence result is corresponding, the system determines that the currently identified temperature characteristic region is valid. Once the temperature characteristic region is determined to be valid, the system can generate a pre-judgment time of the material state change reaching the refractive index sensor according to a preset flow rate model and the pipe length from the temperature probe to the refractive index sensor, combined with the actual propagation delay of the material of this type in the historical data. For example, if the flow rate is known to be 0.5 meters per second and the pipe length is 10 meters, the preliminary propagation time is 20 seconds, and combined with the historical data correction, the final pre-judgment time is obtained.

[0096] Optionally, in combination with Figure 3 As shown, S31 separates the trend signal and the detail signal based on the temperature variation of the temperature characteristic region, and the steps include:

[0097] S311, obtaining a working frequency sequence of the conveying device corresponding in time to the temperature variation of the temperature characteristic region;

[0098] S312, determining a reference frequency based on the working frequency sequence;

[0099] S313, performing time-varying processing on the temperature variation according to the reference frequency to separate the trend signal and the detail signal.

[0100] The working frequency sequence refers to the record of the change of the running frequency of the conveying equipment over time, which can be the rotational speed, vibration frequency or power frequency of the equipment directly measured by a sensor, or the set or actual running frequency data read from the control system of the equipment, and the purpose is to capture the dynamic change of the running state of the conveying equipment, and to provide a basis for subsequent identification and elimination of the interference of the temperature signal. The reference frequency refers to one or a group of representative frequency values extracted from the obtained working frequency sequence, which can be the average value, median value, or dominant frequency component determined by frequency domain analysis, and the purpose is to provide a reference for time-varying processing, so that the processing process can identify and process the temperature signal component related to the working frequency of the equipment. The time-varying processing refers to a signal processing method, the processing parameters or algorithm of which will be dynamically adjusted according to the characteristics of the signal or external reference information over time, which can be realized by adaptive filtering, wavelet transform, or Kalman filter-based algorithm, etc. These algorithms can dynamically adjust the filtering characteristics or decomposition basis according to the reference frequency, so as to finely decompose the temperature change signal at different time points, and the purpose is to effectively distinguish the periodic or quasi-periodic interference caused by the working frequency of the conveying equipment from the real, non-equipment working frequency related temperature change, so as to more accurately separate the trend signal and the detail signal.

[0101] In some preferred embodiments, the specific process of separating the trend signal and the detail signal based on the temperature variation of the temperature characteristic region can be implemented as follows. First, in order to obtain the working frequency sequence of the conveying equipment corresponding to the temperature variation of the temperature characteristic region in time, a speed sensor or a vibration sensor can be deployed on the conveying equipment (such as a pump or a stirrer) to collect its running data in real time. These sensors can output an electrical signal proportional to the device speed or vibration frequency, which is digitized by a data acquisition system at a preset sampling frequency, thereby forming time series data. In addition, the set or actual running frequency parameters of the conveying equipment can also be directly read from its PLC (Programmable Logic Controller) or DCS (Distributed Control System), and time-stamped aligned with the temperature distribution signal collected by the temperature probe to ensure their correspondence in time. Further, based on the obtained working frequency sequence, the reference frequency is determined. For example, the working frequency sequence can be processed by sliding window averaging to calculate the average working frequency in each time window as the reference frequency. Alternatively, the working frequency sequence can be subjected to Fourier transform to identify its main frequency components, and the frequency with the highest energy is taken as the reference frequency. In some cases, if the working frequency of the conveying equipment is discrete and has several gears, the frequency corresponding to the current gear of the equipment can be directly taken as the reference frequency. Subsequently, the temperature variation is processed in time according to the determined reference frequency, and the trend signal and the detail signal are separated. Specifically, an adaptive filtering algorithm can be used, such as an adaptive notch filter based on the least mean square (LMS) algorithm. The filter can dynamically adjust its notch frequency according to the real-time determined reference frequency, thereby effectively filtering out or weakening the periodic interference components in the temperature signal related to the working frequency of the conveying equipment. Another implementation is to use wavelet decomposition. By selecting a suitable wavelet basis function and determining the decomposition scale according to the reference frequency, the temperature variation signal can be decomposed into subbands of different frequency components. The subband corresponding to the reference frequency can be considered as the component in the detail signal caused by the device interference, while the low-frequency component constitutes the trend signal, and the high-frequency non-device-related component constitutes the remaining part of the detail signal. In this way, the trend part and the detail part after removing the device interference in the temperature variation can be accurately separated.

[0102] Optionally, in combination with Figure 4 As shown in FIG. 6, the step of generating the pre-judgment moment in step S35 according to the propagation characteristics of the temperature characteristic region includes:

[0103] S351, identifying and determining the first moment when the temperature characteristic region reaches the refractive index sensor based on the first material state change event, and determining the second moment when the solution concentration change detected by the refractive index sensor occurs;

[0104] S352, determine a time deviation according to the first time and the second time;

[0105] S353, identify and generate an initial prediction time according to the propagation characteristic of the confirmed effective temperature characteristic region corresponding to the second material state change event based on the second material state change event;

[0106] S354, correct the initial prediction time based on the time deviation to generate a prediction time of the material state change reaching the refractive index sensor.

[0107] The first material state change event refers to a specific moment or process in the production process that first occurs or is identified by the system, causing a significant change in the properties of the material (such as concentration, temperature), which can be determined by manual triggering, system automatic identification (for example, by monitoring production batch switching signals, raw material addition signals, or sudden changes in specific process parameters), etc. The purpose is to provide a clear, traceable reference point for measuring the time relationship between temperature change and concentration change. The second material state change event refers to a specific moment or process in the production process that needs to be predicted, which occurs subsequently, causing a significant change in the properties of the material, which can be identified in a similar way to the first material state change event, such as by monitoring subsequent batch switching, new raw material addition or process parameter adjustment again, the purpose is to serve as the target event for which accurate prediction time needs to be generated. The first time refers to the time point when the temperature characteristic region actually reaches the refractive index sensor after the first material state change event occurs, which can be determined by monitoring the significant change in the temperature signal detected by the temperature probe upstream of the refractive index sensor, such as when the slope of the temperature signal exceeds the preset threshold or reaches the peak value, the purpose is to mark the specific time when the temperature change reaches the measurement point. The second time refers to the time point when the solution concentration change is actually detected by the refractive index sensor after the first material state change event occurs, which can be determined by monitoring the significant change in the refractive index signal output by the refractive index sensor, such as when the slope of the refractive index signal exceeds the preset threshold or reaches a stable change, the purpose is to mark the specific time when the concentration change actually reaches the measurement point. The time deviation refers to the time difference between the first time and the second time, that is, the delay between the time when the temperature characteristic region reaches the refractive index sensor and the time when the solution concentration change is actually detected by the refractive index sensor, which can be obtained by directly calculating the second time minus the first time, the purpose is to quantify the actual time lag between temperature change and concentration change when reaching the same measurement point, providing a basis for subsequent prediction correction. Confirming the propagation characteristics of the effective temperature characteristic region refers to the temperature characteristic region that has been verified and determined to actually reflect the change in the state of the material, its dynamic behavior such as propagation speed and attenuation law in the flow path, which can be determined by methods such as historical data analysis, fluid mechanics model calculation or real-time signal processing (for example, by trend verification and frequency correspondence judgment), the purpose is to ensure that the temperature information used to generate the initial prediction time is reliable, avoiding prediction errors caused by invalid temperature signals.The initial prediction time refers to the time when the material state change reaches the refractive index sensor, which is preliminarily estimated according to the propagation characteristics of the confirmed effective temperature characteristic region without considering the actual time delay between the temperature change and the concentration change. It can be calculated based on the starting time, propagation distance and propagation speed of the temperature characteristic region, and its purpose is to provide a preliminary prediction value based on temperature propagation as the starting point for subsequent correction. Correction refers to adjusting the initial prediction time to make it closer to the actual time when the material state change reaches the refractive index sensor. It can be adjusted by directly adding or subtracting the time deviation to the initial prediction time, or by a more complex model (for example, a correction model established based on historical data), and its purpose is to eliminate or reduce the prediction error caused by the time delay between the temperature change and the concentration change, and to improve the accuracy of the prediction.

[0108] In some preferred embodiments, the application is implemented as follows. For example, in a continuous flow chemical reactor production line, when the raw material batch is switched, it can be identified as a material state change event. The system first identifies and determines the first time when the temperature characteristic region reaches the refractive index sensor based on the first batch switching event. This can be accurately recorded by monitoring the significant change in the temperature signal detected by the temperature probe upstream of the refractive index sensor (for example, the slope of the temperature curve exceeds the preset threshold). At the same time, the system determines the second time when the concentration change of the solution detected by the refractive index sensor occurs, which can be recorded by monitoring the significant change in the refractive index signal output by the refractive index sensor (for example, the step or trend change of the refractive index curve). Then, the system calculates and determines the time deviation between the first time and the second time according to the determined first time and second time. For example, if the temperature change reaches the sensor 5 seconds earlier than the concentration change, the time deviation is 5 seconds. When the second batch switching event occurs subsequently, the system identifies it as a second material state change event. At this time, the system generates an initial prediction time according to the propagation characteristics of the confirmed effective temperature characteristic region corresponding to the second material state change event. For example, the system can preliminarily calculate the time when the material state change reaches the sensor according to the distance and estimated propagation speed of the temperature characteristic region propagating from upstream to the refractive index sensor, assuming that it is 10 seconds after the temperature characteristic region arrives. In order to improve the accuracy of the prediction, the system will correct the initial prediction time based on the previously determined time deviation (for example, 5 seconds). Specifically, if the time deviation indicates that the concentration change lags behind the temperature change, the system will add the time deviation to the initial prediction time to generate the final prediction time when the material state change reaches the refractive index sensor. For example, the corrected prediction time will be 10 seconds plus 5 seconds, i.e. 15 seconds.

[0109] Optionally, in combination with Figure 5According to the propagation characteristics of the temperature characteristic region in step S35, the step of generating the pre-judgment time point includes:

[0110] A1, obtaining historical material state change events; each historical material state change event includes a historical time deviation and a historical process parameter;

[0111] A2, establishing a correspondence between the process parameters and the time deviation based on the historical material state change events;

[0112] A3, for the second material state change event, obtaining the current process parameter corresponding to the second material state change event;

[0113] A4, according to the correspondence between the process parameters and the time deviation and the current process parameter, determining the verified time deviation corresponding to the second material state change event;

[0114] A5, correcting the initial pre-judgment time point based on the verified time deviation to generate the pre-judgment time point of the material state change reaching the refractive index sensor.

[0115] Wherein, the historical material state change event refers to an event that occurs at a certain time point in the past and causes the material state to change, which can include reactant addition, temperature mutation, stirring speed adjustment, etc., and its purpose is to collect data samples for subsequent analysis; the historical time deviation refers to the time delay required for the material state change to propagate from its source to the refractive index sensor and be detected by it when the historical material state change event occurs, and its purpose is to quantify the dynamic characteristics of material propagation; the historical process parameter refers to various production operation conditions that affect the material propagation characteristics and the time deviation when the historical material state change event occurs, which can include reaction temperature, feed rate, stirring rate, material viscosity, pipeline pressure, etc., and its purpose is to record external factors that affect the time deviation; the correspondence between the process parameters and the time deviation refers to a mathematical model or mapping rule that describes how the time deviation changes under different process parameter conditions, which can be established by regression analysis, machine learning model, lookup table or piecewise function, etc., and its purpose is to reveal the influence law of process parameters on the time deviation; the current process parameter refers to the process operation condition related to the event that is monitored or recorded by the system in real time when the second material state change event occurs, and its purpose is to provide input for predicting the current time deviation; the verified time deviation refers to the time value that more accurately reflects the material propagation delay under the current working condition after the initial time deviation is corrected according to the current process parameter and the established correspondence, and its purpose is to provide a more accurate time deviation that is corrected by the influence of the process parameter.

[0116] In some preferred embodiments, specifically, the system can continuously collect material state change event data during the production process. For example, when a new batch of material is fed or a process parameter is adjusted in the reaction kettle, the first time when the temperature characteristic region reaches the refractive index sensor and the second time when the solution concentration changes can be recorded, so as to calculate the historical time deviation. At the same time, the corresponding process parameters at that time, such as reaction temperature, stirring speed, feed flow rate, etc., can be recorded, which can be stored in the database to form a record of historical material state change events. Based on these historical material state change events, a multiple linear regression model or a support vector regression model can be used to establish the corresponding relationship between the process parameters and the time deviation. For example, the historical time deviation can be used as the dependent variable, and the historical process parameters can be used as the independent variable, and the functional relationship between them can be learned by training the model. Alternatively, the process parameter space can be divided into multiple intervals, and the average time deviation of the historical data in each interval can be calculated to form a lookup table. When a new second material state change event occurs, for example, when the system detects a new temperature characteristic region and generates an initial prediction time, the current reaction temperature, feed flow rate and other process parameters can be obtained from the distributed control system or programmable logic controller in real time. Subsequently, the current process parameters obtained are input into the regression model or lookup table established before. If a regression model is used, a calibrated time deviation is obtained by the model prediction; if a lookup table is used, the corresponding calibrated time deviation is found according to the interval in which the current process parameters are located. Finally, this calibrated time deviation can be applied to the correction of the initial prediction time. For example, if the initial prediction time is T_initial and the calibrated time deviation is Delta_T_calibrated, then the final prediction time of the material state change reaching the refractive index sensor is T_initial plus Delta_T_calibrated.

[0117] Optionally, in combination with Figure 6 As shown in FIG. 2, the step of A2 of establishing the corresponding relationship between the process parameters and the time deviation based on the historical material state change events comprises:

[0118] A21, based on the historical material state change events, dividing the numerical range of the historical process parameters in the historical material state change events into a plurality of process parameter intervals;

[0119] A22, for each process parameter interval, obtaining the historical material state change events corresponding to the historical process parameters falling into the process parameter interval from the historical material state change events, denoted as historical control events;

[0120] A23, calculating the time deviation representative value based on the historical time deviations of the historical control events;

[0121] A24, mapping between the process parameter interval and the time deviation representative value as a corresponding relationship.

[0122] Wherein, the process parameter interval refers to a plurality of sub-ranges obtained by discretizing a continuous historical process parameter value range, which can be achieved by equal-width division, equal-frequency division or dynamic division based on clustering analysis, aiming to classify similar process conditions to reduce the influence of single data point outliers; the historical reference event refers to a set of events whose historical process parameters fall into the interval, selected from all historical material state change events within a specific process parameter interval, which can be obtained by database query, data filtering or index lookup, aiming to ensure that the data set used to calculate the time deviation representative value has a similar process background; the time deviation representative value refers to a single value obtained by comprehensive statistical calculation of the historical time deviation of all historical reference events within a specific process parameter interval, which can be calculated by average, median, weighted average or center value based on data distribution characteristics, aiming to eliminate random errors and obtain a more representative time deviation; the mapping refers to a kind of association between the process parameter interval and the time deviation representative value, as a corresponding relationship, which can be achieved by lookup table, piecewise function, decision tree or neural network model, aiming to provide accurate basis for subsequent prediction time correction.

[0123] In some preferred embodiments, the process of establishing the correspondence between the process parameter and the time deviation can be implemented as follows. First, the system can obtain a large amount of historical material state change event data, which are usually stored in a database, each event records its corresponding historical time deviation and historical process parameter. For example, if the historical process parameter is the reaction temperature, its value range can be from 50 degrees Celsius to 100 degrees Celsius. The system can divide this temperature range into several equal-width process parameter intervals, for example, every 5 degrees Celsius as an interval, forming 50-55℃, 55-60℃, and so on. Then, for each such process parameter interval, the system can query and filter out all historical material state change events whose historical process parameters fall into the interval from the database. For example, for the temperature interval of 50-55℃, the system will find all historical material state change events that occur between 50℃ and 55℃, and collect these events as historical control events. Then, the system can calculate the representative value of the time deviation for this process parameter interval based on the historical time deviations contained in the historical control events. For example, the arithmetic mean of these historical time deviations can be calculated as the representative value, or in order to deal with possible outliers, the median or trimmed mean can also be calculated. If the historical time deviations of the historical control events exhibit a specific data distribution characteristic, such as normal distribution or skew distribution, the system can also determine the data concentration region according to these distribution characteristics, and calculate the representative value based on the historical time deviations within the region, for example, calculate the average value within the peak region. Finally, the system can establish a lookup table or a piecewise function to associate each process parameter interval with its calculated time deviation representative value, thereby forming the correspondence between the process parameter and the time deviation. When a new second material state change event occurs, the system can obtain its current process parameter, and then quickly and accurately find the corresponding verified time deviation by querying the lookup table or calculating the piecewise function, and then use it to correct the initial prediction time.

[0124] Optionally, in combination with Figure 7 As shown, the step of A23 of calculating the representative value of the time deviation based on the historical time deviations of the historical control events includes:

[0125] A231, determining the data distribution characteristic of the historical time deviations of the historical control events;

[0126] A232, determining the data concentration region according to the data distribution characteristic;

[0127] A233, calculating the representative value of the time deviation based on the historical time deviations within the data concentration region.

[0128] The data distribution characteristics of the historical time deviation of the historical control event are statistical characteristics of the data distribution on the numerical axis, and the statistical characteristics include concentration tendency, dispersion degree, skewness, kurtosis, etc. The statistical parameters of the data distribution can be obtained by statistical analysis of the historical time deviation, or the data distribution can be observed visually by drawing a histogram or a kernel density estimation graph, or the characteristics of the data distribution can be quantified by fitting a probability distribution model. The purpose is to identify the concentration tendency, dispersion degree and possible outliers in the data, and to provide a basis for subsequent data screening and representative value calculation.

[0129] The determination of the data concentration region is to identify the range with high numerical density and high frequency of occurrence in the historical time deviation data. The peak region can be determined based on the density distribution of the historical time deviation, or the main concentration region of the data can be identified and determined by statistical methods such as quartile range method, Z-score method, clustering analysis, etc. The purpose is to exclude the interference of abnormal values or noise data, focus on the main distribution region of the data, and improve the reliability of the representative value.

[0130] The calculation of the time deviation representative value is to calculate a single numerical value that can effectively represent the overall trend or central position of the historical time deviation in a specific process parameter interval using high-quality data after screening. The mean, median, weighted average, truncated mean or mode of the historical time deviation in the data concentration region can be calculated. The purpose is to improve the accuracy of the representative value.

[0131] In some preferred embodiments, the application is implemented as follows: when calculating the time deviation representative value, first, the historical time deviation of the historical control event can be statistically analyzed, such as calculating the mean, standard deviation, skewness and kurtosis, and drawing a histogram or kernel density estimation graph to visually observe the data distribution characteristics. For example, if it is observed that the historical time deviation is mainly concentrated in a certain range and the distribution is close to normal, but there are a small number of extreme values.

[0132] Secondly, according to these data distribution characteristics, the data concentration region can be determined. For example, the interquartile range (IQR) method can be used to identify and exclude outliers, i.e. data points falling below the first quartile (Q1) minus 1.5 times IQR or above the third quartile (Q3) plus 1.5 times IQR are considered outliers and are excluded, thereby determining a concentration region containing most of the normal data. Alternatively, the peak region of the historical time deviation distribution can be identified by density estimation, and the peak region is taken as the data concentration region.

[0133] Finally, based on the determined historical time deviation of the historical control event, a representative value of the time deviation is calculated. For example, the arithmetic mean of all historical time deviations in the set region can be calculated as the representative value of the time deviation, or if the data distribution is slightly skewed, the median can be used as the representative value to better reflect the central tendency of the data, so as to obtain an optimized and more representative time deviation value.

[0134] Optionally, the step of determining the data distribution characteristics of the historical time deviation of the historical control event by A231 comprises:

[0135] The statistical analysis of the historical time deviation of the historical control event is performed to obtain statistical parameters of the data distribution as the data distribution characteristics.

[0136] The statistical analysis refers to the process of systematic collection, arrangement, description, analysis and inference of the data set, aiming to reveal the inherent rules, trends and characteristics of the data, which can be achieved by descriptive statistical methods or inferential statistical methods. The statistical parameters of the data distribution refer to numerical indicators for quantitatively describing the characteristics of the data distribution, such as shape, central tendency, dispersion, symmetry and kurtosis, which can include but are not limited to mean, median, mode, standard deviation, variance, skewness, kurtosis, quartile, etc. The data distribution characteristics refer to the overall rules and properties of the data set revealed by statistical analysis, which can reflect the distribution shape of the data on the numerical axis, and can be represented as a set of statistical parameters or described by a mathematical model such as probability density function or cumulative distribution function.

[0137] In some preferred embodiments, specifically, in order to determine the data distribution characteristics of the historical time deviation of the historical control event, a large number of historical control events can be collected first, and the corresponding historical time deviation values of each event are recorded. For example, hundreds or even thousands of historical time deviation data points can be collected. Then, statistical analysis of these historical time deviation data is performed using data processing software or programming language. The specific operation can include: calculating the arithmetic mean of the data to understand the central position; calculating the standard deviation or variance to measure the dispersion of the data; calculating the median to reflect the intermediate value of the data; calculating the mode to find the value with high frequency; and calculating the skewness to evaluate the symmetry of the data distribution, and the kurtosis to evaluate the sharpness of the data distribution. These calculated statistical quantities, such as mean, standard deviation, median, mode, skewness and kurtosis, are collectively used as the data distribution characteristics of the historical time deviation data set. The set of these statistical parameters can comprehensively and quantitatively describe the overall distribution of the historical time deviation, providing a solid data foundation for determining the set region in the subsequent steps.

[0138] Optionally, in combination with Figure 8As shown, the step of A232 of determining the data-centralized region according to the data distribution characteristics includes:

[0139] A2321, determining the density distribution of the historical time deviation according to the data distribution characteristics;

[0140] A2322, determining the peak region according to the density distribution;

[0141] A2323, taking the peak region as the data-centralized region.

[0142] Wherein, the data distribution characteristics refer to the properties describing the scattering of a group of data on the numerical axis, which can be represented by statistical parameters such as mean, variance, skewness or kurtosis, or by non-parametric methods such as histogram or kernel density estimation, the purpose of which is to provide basic information for subsequent analysis of data-centralized trends. The density distribution refers to the density or probability distribution of data in different value ranges, which can be obtained by kernel density estimation (KDE) method, histogram smoothing processing or parameterized probability density function fitting, etc., the purpose of which is to intuitively show the concentration and mode of data on the numerical axis. The peak region refers to the continuous interval with the highest local or global density on the density distribution curve, which can be determined by identifying the maximum value point on the density distribution curve, and expanding to the range where the density significantly decreases or reaches the preset threshold on both sides of the maximum value point, the purpose of which is to accurately locate the core region of the most concentrated data. The data-centralized region refers to the range in which the historical time deviation data is highly concentrated in numerical value, which can be represented by the peak region, the purpose of which is to provide a highly representative data subset for calculating the representative value of the time deviation.

[0143] In some preferred embodiments, the present application is implemented as follows: assuming that a group of historical time deviation data of historical control events has been obtained, for example, by performing preliminary statistical analysis on these historical time deviation data, the mean, standard deviation and other data distribution characteristics can be obtained. In order to more accurately determine the data-centralized region, the kernel density estimation (KDE) method can be used to determine the density distribution of the historical time deviation. Specifically, a Gaussian kernel function can be assigned to each historical time deviation data point, and all kernel functions are superimposed to obtain a smooth density distribution curve. This curve can intuitively show the probability density of historical time deviation at different numerical values.

[0144] Subsequently, according to the density distribution curve, a peak region can be identified and determined. For example, by finding the local maximum points on the density distribution curve, and extending to both sides of the maximum points to the point where the density value drops to a certain percentage (e.g. 50%) of the peak density or the inflection point where the second derivative of the curve is zero, a continuous peak region can be defined. If there are multiple peaks, according to a preset rule, for example, selecting the region corresponding to the highest peak, or merging all significant peak regions.

[0145] Finally, the determined peak region is taken as the data concentration region of the historical time deviation. For example, if the density distribution curve presents a significant peak between 10 seconds and 15 seconds, then the interval of 10 seconds to 15 seconds is determined as the data concentration region. In this way, when calculating the representative value of the time deviation later, only the data in this concentration region can be used for calculation, so as to avoid the influence of outliers or atypical data on the accuracy of the representative value.

[0146] A refractive index-based solution concentration online monitoring system for performing refractive index-based solution concentration online monitoring, comprising Figure 9 As shown in the figure, the refractive index-based solution concentration online monitoring system 1 comprises:

[0147] A signal acquisition module 11 for acquiring the refractive index signal of the refractive index sensor arranged along the continuous flow path in the reaction kettle, the physical disturbance signal of the physical disturbance sensor, and the temperature distribution signal of the temperature probe arranged upstream of the refractive index sensor;

[0148] An interference identification module 12 for identifying the physical interference component in the refractive index signal according to the time correspondence relationship between the physical disturbance signal and the refractive index signal;

[0149] A pre-judgment time generation module 13 for identifying the temperature characteristic region propagating along the continuous flow path in the reaction kettle based on the temperature distribution signal, and generating the pre-judgment time of the material state change reaching the refractive index sensor according to the temperature characteristic region;

[0150] A change recording module 14 for monitoring the refractive index signal based on the pre-judgment time, and recording the current change as the refractive index change caused by the solution concentration change when the refractive index signal changes corresponding to the pre-judgment time;

[0151] A concentration monitoring module 15 for temperature compensating the refractive index change using the temperature distribution signal to obtain information representing the solution concentration.

[0152] Among them, the signal acquisition module refers to the collection of raw data from different types of sensors, which can be an integrated data acquisition unit, and its purpose is to provide comprehensive input for subsequent data processing and analysis; the interference identification module refers to separating the interference caused by non-concentration factors from the original refractive index signal, which can be a signal processing unit, and its purpose is to ensure the purity and accuracy of the refractive index data; the pre-judgment time generation module refers to predicting the time when the material state changes to the monitoring point according to the temperature change trend, which can be a prediction analysis unit, and its purpose is to improve the timeliness of the system's response to concentration changes; the change recording module refers to verifying and recording the refractive index signal at a specific pre-judgment time, which can be a data verification and storage unit, and its purpose is to accurately distinguish between real concentration changes and external interference; the concentration monitoring module refers to temperature correction of the refractive index data and calculation of the solution concentration, which can be a data processing and output unit, and its purpose is to provide accurate solution concentration information.

[0153] In some preferred embodiments, the system can be implemented as follows. The signal acquisition module can include multiple analog-to-digital converters and data acquisition cards, which are connected to the fiber refractometer as the refractive index sensor, the ultrasonic sensor or the piezoelectric sensor as the physical disturbance sensor, and the thermocouple array or the thermal resistance as the temperature probe to realize the synchronous acquisition of multi-source data. The interference identification module can be realized by an embedded processor or software algorithm on an industrial personal computer, which can perform signal processing techniques such as wavelet transform or Fourier transform, combined with threshold judgment or machine learning models to identify and filter out physical disturbances caused by bubbles or solid particles. The pre-judgment time generation module can be realized by an independent microcontroller or a shared processing unit with the interference identification module, and the algorithm running inside can analyze the timing characteristics of the temperature distribution signal, such as determining the propagation speed and arrival time of the temperature characteristic region through curve fitting or pattern recognition. The change recording module can be a data storage and management unit, such as a non-volatile memory or a database, controlled by the host unit, triggering data recording at the pre-judgment time and marking it as a concentration change event. The concentration monitoring module can be the core processing part of the host unit, which executes temperature compensation algorithms such as based on a pre-set refractive index-temperature-concentration lookup table or mathematical model, and outputs the final concentration information through a human-machine interface or a communication interface, such as through Modbus or Ethernet / IP protocol to the upper computer or control system.

[0154] Through the technical scheme, the system provides a specific implementation carrier capable of effectively executing the solution concentration online monitoring method based on the refractive index. The system can apply complex data processing logic and algorithms to an actual industrial production environment by integrating function modules such as signal acquisition, interference identification, pre-judgment time generation, change recording, and concentration monitoring. This makes the original method no longer stay at the theoretical level, but can be efficiently and stably deployed and run, thereby overcoming the limitation that only relying on the method flow cannot realize the solution concentration online monitoring. Finally, the system ensures that under complex working conditions with adverse factors such as delivery delay, temperature change, and physical disturbance, the solution concentration change can still be quickly captured, accurately identified, and accurately fed back, thereby improving the reliability and practicality of online monitoring.

[0155] The above merely describes the embodiments of the present application and is not used to limit the protection scope of the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for on-line monitoring of concentration of a solution based on refractive index, characterized in that, The method comprises the following steps: Obtaining the refractive index signal of the refractive index sensor arranged along the continuous flow path in the reaction kettle, the physical disturbance signal of the physical disturbance sensor, and the temperature distribution signal of the temperature probe arranged upstream of the refractive index sensor; According to the time correspondence of the physical disturbance signal and the refractive index signal, the physical disturbance component in the refractive index signal is identified; Based on the temperature distribution signal, a temperature characteristic region propagating along the continuous flow path in the reaction kettle is identified, and a pre-judgment time when the material state change reaches the refractive index sensor is generated according to the temperature characteristic region; Based on the pre-judgment time, the refractive index signal is monitored, and when the refractive index signal changes corresponding to the pre-judgment time, the current change is recorded as the refractive index change caused by the change of the solution concentration; The temperature compensation is performed on the refractive index change by using the temperature distribution signal to obtain the information representing the solution concentration.

2. The method according to claim 1, wherein, The step of identifying the temperature characteristic region propagating along the continuous flow path in the reaction kettle based on the temperature distribution signal and generating the pre-judgment time when the material state change reaches the refractive index sensor according to the temperature characteristic region comprises the following steps: Based on the temperature change of the temperature characteristic region, the trend signal and the detail signal are separated; The trend signal is verified based on the preset change criterion to obtain a trend verification result; The frequency domain analysis is performed on the detail signal to determine the characteristic frequency of the detail signal; It is judged whether the characteristic frequency corresponds to the working frequency of the conveying device of the driving fluid to obtain a frequency correspondence result; When the trend verification result is passed and the frequency correspondence result is corresponding, the temperature characteristic region is determined as valid, and the pre-judgment time is generated according to the propagation characteristics of the temperature characteristic region.

3. The method according to claim 2, wherein, The step of separating the trend signal and the detail signal based on the temperature change of the temperature characteristic region comprises the following steps: Obtaining the working frequency sequence of the conveying device corresponding to the temperature change of the temperature characteristic region in time; Based on the working frequency sequence, a reference frequency is determined; According to the reference frequency, the temperature change is time-varying processed to separate the trend signal and the detail signal.

4. The method according to claim 2, wherein, The step of generating the pre-judgment time according to the propagation characteristics of the temperature characteristic region comprises the following steps: Identifying and determining the first time when the temperature characteristic region reaches the refractive index sensor based on the first material state change event, and determining the second time when the solution concentration change detected by the refractive index sensor occurs; According to the first time and the second time, a time deviation is determined; Identifying and generating an initial pre-judgment time according to the propagation characteristics of the temperature characteristic region corresponding to the second material state change event based on the second material state change event; The initial pre-judgment time is corrected based on the time deviation to generate the pre-judgment time when the material state change reaches the refractive index sensor.

5. The method according to claim 4, wherein the method is characterized by, The step of generating the pre-judgment time according to the propagation characteristics of the temperature characteristic region comprises the following steps: Obtaining historical material state change events; each historical material state change event comprises a historical time deviation and a historical process parameter; Based on the historical material state change events, a corresponding relationship between the process parameters and the time deviations is established; For the second material state change event, a current process parameter corresponding to the second material state change event is obtained; According to the correspondence between the process parameter and the time deviation and the current process parameter, a verified time deviation corresponding to the second material state change event is determined; Based on the verified time deviation, the initial prediction time is corrected to generate a prediction time of the material state change reaching the refractive index sensor.

6. The method according to claim 5, wherein the step of calculating the concentration of the solution is performed by using the following equation: ###0001### wherein, n is the refractive index of the solution, n0 is the refractive index of the solvent, and c is the concentration of the solution. The step of establishing the correspondence between the process parameter and the time deviation based on the historical material state change event comprises: Based on the historical material state change event, the numerical range of the historical process parameter in the historical material state change event is divided into several process parameter intervals; For each process parameter interval, the historical material state change events corresponding to the historical process parameters falling into the process parameter interval are obtained from the historical material state change events, which are recorded as historical control events; Based on the historical time deviation of the historical control event, a time deviation representative value is calculated; The mapping between the process parameter interval and the time deviation representative value is established as the correspondence.

7. The method according to claim 6, wherein, The step of calculating the time deviation representative value based on the historical time deviation of the historical control event comprises: Determine the data distribution characteristics of the historical time deviation of the historical control event; According to the data distribution characteristics, a data concentration area is determined; Based on the historical time deviation in the data concentration area, a time deviation representative value is calculated.

8. The method according to claim 7, wherein the step of calculating the concentration of the solution is performed by using the following equation: ###0002### wherein, n is the refractive index of the solution, n0 is the refractive index of the solvent, and c is the concentration of the solution. The step of determining the data distribution characteristics of the historical time deviation of the historical control event comprises: Statistical analysis is performed on the historical time deviation of the historical control event to obtain statistical parameters of data distribution as data distribution characteristics.

9. The method according to claim 7, wherein the step of calculating the concentration of the solution is performed by using the following equation: ###0002### wherein, n is the refractive index of the solution, n0 is the refractive index of the solvent, c is the concentration of the solution, and K is a constant. The step of determining the data concentration area according to the data distribution characteristics comprises: Based on the data distribution characteristics, the density distribution of the historical time deviation is determined; According to the density distribution, a peak value area is determined; The peak value area is taken as the data concentration area.

10. A refractive index-based solution concentration on-line monitoring system for performing refractive index-based solution concentration on-line monitoring, characterized by, Comprise: The signal acquisition module is used for acquiring the refractive index signal of the refractive index sensor arranged along the continuous flow path in the reaction kettle, the physical disturbance signal of the physical disturbance sensor, and the temperature distribution signal of the temperature probe arranged upstream of the refractive index sensor; The interference identification module is used for identifying the physical interference component in the refractive index signal according to the time correspondence between the physical disturbance signal and the refractive index signal; The prediction time generation module is used for identifying the temperature characteristic area propagating along the continuous flow path in the reaction kettle based on the temperature distribution signal, and generating the prediction time of the material state change reaching the refractive index sensor according to the temperature characteristic area; The change recording module is used for monitoring the refractive index signal based on the prediction time, and recording the current change as the refractive index change caused by the solution concentration change when the refractive index signal changes corresponding to the prediction time; The concentration monitoring module is used for temperature compensation of the refractive index change by using the temperature distribution signal to obtain information representing the solution concentration.

Images