Intelligent Monitoring Method for the Production Process of Waterborne Coatings Based on Data Analysis

Abstract

The present invention relates to the technical field of intelligent control of production processes, and particularly to an intelligent monitoring method for the production process of waterborne coatings based on data analysis. The method includes obtaining temperature data and concentration data; determining a temperature change period segment according to the temperature data; determining a temperature data deviation index according to the temperature fluctuations in the temperature change period segment; determining a chemical reaction rate according to the change of the concentration data; dividing into two time categories of the pre-reaction period and the post-reaction period, clustering the chemical reaction rates under the time categories to obtain time clusters; determining a chemical reaction deviation index at the sampling moment in the time clusters; combining the temperature data deviation index and the chemical reaction deviation index to determine the necessity of temperature regulation; and performing temperature PID control adjustment in the reaction kettle according to the necessity of temperature regulation. The present invention can perform timely temperature control, ensure that the chemical reaction is always in the best state, and guarantee the temperature consistency of the entire production process.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial control, and particularly relates to an intelligent monitoring method for the production process of waterborne coatings based on data analysis. Background Art

[0002] Waterborne coatings are coatings that use water as the main solvent or dispersion medium, are environmentally friendly and low in pollution, and are widely used in the construction and industrial fields. The production process usually includes steps such as batching, mixing, dispersing, paint adjustment, filtering, and packaging, and a stable waterborne coating system is formed through physical mixing and chemical reactions. The chemical reactions involved in waterborne coatings are very sensitive to temperature. Too high or too low temperature may cause the reaction rate to be unstable, thereby affecting the performance of the coatings.

[0003] In the prior art, temperature data is usually obtained through temperature sensors, and PID control adjustment is performed according to the temperature data to ensure temperature stability. In this way, due to the continuous change of the reaction rate during the production process of waterborne coatings, the temperature fluctuation is relatively intense, and the temperature change - PID control regulation process has hysteresis, and the PID control adjustment response is slow. As a result, when large fluctuations occur, the temperature is abnormal but the PID controller still does not respond to the adjustment, making the overall regulation untimely and the temperature consistency of the production process poor. Summary of the Invention

[0004] In order to solve the technical problem that in the related art, the PID control adjustment according to temperature data has a slow response, resulting in abnormal temperature but the PID controller still not responding to the adjustment when large fluctuations occur, making the overall regulation untimely and the temperature consistency of the production process poor, the present invention provides an intelligent monitoring method for the production process of waterborne coatings based on data analysis. The specific technical solutions adopted are as follows:

[0005] The present invention proposes an intelligent monitoring method for the production process of waterborne coatings based on data analysis. The method includes:

[0006] Periodically obtain the temperature data and concentration data of the materials in the reaction kettle at different sampling times; segment the time according to the extreme values of the temperature data to determine different temperature change cycle segments;

[0007] Determine the temperature cycle characteristic index of each temperature change cycle segment according to the temperature fluctuation at each sampling time in different temperature change cycle segments; determine the temperature data deviation index of each temperature change cycle segment according to the temperature fluctuation comparison and temperature cycle characteristic index difference between any temperature change cycle segment and all other temperature change cycles;

[0008] Determine the chemical reaction rate at each sampling moment according to the change of concentration data at adjacent sampling moments; divide the sampling moments into two time categories, namely the pre-reaction period and the post-reaction period, according to the peak value of the temperature data, and cluster the values of the chemical reaction rates at different sampling moments in each time category to obtain time clusters; determine the chemical reaction deviation index at each sampling moment according to the difference in the chemical reaction rates between any sampling moment and other sampling moments in its time cluster.

[0009] Combine the temperature data deviation index and the chemical reaction deviation index to determine the temperature adjustment degree at each sampling moment; determine the necessity of temperature control at each sampling moment according to the change of the temperature adjustment degree within the preset neighborhood time sequence range of each sampling moment.

[0010] Perform temperature PID control adjustment in the reaction kettle according to the necessity of temperature control at the current moment.

[0011] Further, perform time segmentation according to the extreme value of the temperature data to determine different temperature change period segments, including:

[0012] Take the time range between the minima of the two closest temperature data as a temperature change period segment.

[0013] Further, determine the temperature period characteristic index of each temperature change period segment according to the temperature fluctuation at each sampling moment in different temperature change period segments, including:

[0014] Take the range of the temperature data of all sampling moments in the same temperature change period segment as the first period index;

[0015] Take the time interval from the start moment of the temperature change period segment to any sampling moment as the target interval corresponding to the sampling moment; calculate the ratio of the target interval to the duration of the temperature change period segment as the second period index corresponding to the sampling moment.

[0016] Determine the standard period coefficient of the temperature change period segment according to the second period index and the first period index of all sampling moments in the same temperature change period segment. The corresponding calculation formula is:

[0017] ; where represents the standard period coefficient of the i-th temperature change period segment, represents the first period index of the i-th temperature change period segment, represents the second period index of the j-th sampling moment in the i-th temperature change period segment, represents the pi, K represents the number of all sampling moments in the i-th temperature change period segment;

[0018] Calculate the absolute value of the difference between the mean of the temperature data at all sampling moments in the temperature change period segment and the standard period coefficient as the standard temperature difference;

[0019] Perform min-max normalization on the opposite number of the standard temperature difference to obtain the temperature period characteristic index of the temperature change period segment.

[0020] Furthermore, determine the temperature data deviation index of each temperature change period segment according to the temperature fluctuation comparison and the difference in temperature period characteristic indexes between any temperature change period segment and all other temperature change periods, including:

[0021] Calculate the DTW value of the temperature values within any two temperature change period segments based on the dynamic time warping algorithm, and perform min-max normalization on the DTW value to obtain the first deviation coefficient of the temperature fluctuation within the two temperature change period segments;

[0022] Take the absolute value of the difference between the temperature period characteristic indexes of any two temperature change period segments as the second deviation coefficient of the temperature fluctuation within the two temperature change period segments;

[0023] Calculate the product of the first deviation coefficient and the second deviation coefficient, and perform min-max normalization to obtain the period deviation index of the two temperature change period segments;

[0024] Take the mean of the period deviation indexes of any temperature change period segment and all other temperature change period segments as the temperature data deviation index corresponding to any temperature change period segment.

[0025] Furthermore, determine the chemical reaction rate at each sampling moment according to the change in the concentration data at adjacent sampling moments, including:

[0026] Take any sampling moment as the target moment, and determine the two sampling moments closest to the target moment as the moments to be analyzed;

[0027] Take the mean of the absolute values of the differences between the concentration data at the target moment and the concentration data at each moment to be analyzed as the chemical reaction rate at the target moment, and adjust the target moment to obtain the chemical reaction rate at each sampling moment.

[0028] Furthermore, cluster the numerical values of the chemical reaction rates at different sampling moments in each time category to obtain time clusters, including:

[0029] Set the value of k to 2, and perform k-means clustering on the numerical values of the chemical reaction rates at all sampling moments in the same time category to obtain 2 clustering clusters in each time category as the time clusters.

[0030] Further, according to the difference in the chemical reaction rate between any sampling moment and other sampling moments in the time cluster to which it belongs, determine the chemical reaction deviation index for each sampling moment, including:

[0031] Take the mean value of the chemical reaction rates of all sampling moments in the time cluster as the first deviation judgment factor, and take the median as the second deviation judgment factor;

[0032] Take the absolute value of the difference between the chemical reaction rate of any sampling moment in the time cluster and the first deviation judgment factor of the time cluster as the first deviation coefficient;

[0033] Take the absolute value of the difference between the chemical reaction rate of any sampling moment in the time cluster and the second deviation judgment factor of the time cluster as the second deviation coefficient;

[0034] Calculate the product of the first deviation coefficient and the second deviation coefficient, and perform maximum-minimum normalization processing as the chemical reaction deviation index of the sampling moment.

[0035] Further, combine the temperature data deviation index and the chemical reaction deviation index to determine the temperature adjustment degree for each sampling moment, including:

[0036] Calculate the product of the chemical reaction deviation index at the same sampling moment and the temperature data deviation index of the temperature change period segment to which the sampling moment belongs, to obtain the temperature adjustment degree corresponding to the sampling moment.

[0037] Further, according to the change in the temperature adjustment degree of each sampling moment within the preset neighborhood time series range, determine the necessity of temperature control for each sampling moment, including:

[0038] Take any sampling moment as the target moment, and determine the two sampling moments before the target moment in the time series as the control influence moments;

[0039] Calculate the mean value of the temperature adjustment degree of the target moment and the temperature adjustment degrees of all control influence moments, and perform maximum-minimum normalization processing to obtain the necessity of temperature control for the target moment.

[0040] Further, according to the necessity of temperature control at the current moment, perform PID control adjustment of the temperature in the reaction kettle, including:

[0041] When the necessity of temperature control is greater than the preset necessity threshold, trigger the PID control adjustment of the temperature in the reaction kettle.

[0042] The present invention has the following beneficial effects:

[0043] The present invention periodically acquires temperature data and concentration data, and then determines the temperature data deviation index for each temperature change period segment according to the numerical fluctuations of the temperature data in different temperature change period segments. The temperature data deviation index characterizes the data anomaly situation of the temperature data in the corresponding temperature change period segment. Then, the chemical reaction rate is determined according to the change of the concentration data, and the chemical reaction deviation index for each sampling moment is determined according to the chemical reaction rate distribution at each sampling moment in the same time cluster. The chemical reaction deviation index characterizes the anomaly situation of the chemical reaction at the corresponding sampling moment. In summary, the temperature adjustment degree is determined by combining the temperature data deviation index and the chemical reaction deviation index, and further the necessity of temperature control is determined. The temperature PID control adjustment in the reaction kettle is carried out according to the necessity of temperature control. By combining the data information of multiple dimensions of temperature and concentration in the reaction kettle, the present invention effectively analyzes the reaction state at each moment, avoids the problem of untimely overall control and slow response of PID control adjustment caused by only adjusting according to the temperature deviation, ensures that the chemical reaction is always in the best state, and guarantees the temperature consistency of the entire production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0045] Figure 1 Flowchart of an intelligent monitoring method for an aqueous coating production process based on data analysis provided by an embodiment of the present invention;

[0046] Figure 2 Schematic diagram of an aqueous coating production line provided by an embodiment of the present invention;

[0047] Figure 3 Schematic diagram of the division of temperature change period segments provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following, in combination with the accompanying drawings and preferred embodiments, details the specific implementation manners, structures, features, and effects of an intelligent monitoring method for an aqueous coating production process based on data analysis proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0049] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the technical field to which this invention belongs.

[0050] To enable those skilled in the art to understand the content of this invention, some basic terms will be explained and elaborated in detail below.

[0051] Waterborne coatings are coatings that use water as a solvent or a dispersion medium. They are a relatively new type of coating in the coating market and include three types: water-soluble coatings, water-dilutable coatings, and water-dispersible coatings (latex coatings).

[0052] A reaction kettle is a device used for chemical reactions, physicochemical processes, and laboratory research. It is usually made of steel plates with a certain thickness and has high corrosion resistance and the ability to withstand high temperature and pressure. It is widely used in the pharmaceutical, chemical, food processing, and other fields to meet the needs of different processes.

[0053] A PID controller (Proportional-Integral-Derivative controller) is a common feedback control method in industrial control applications and consists of a proportional unit P, an integral unit I, and a derivative unit D. The basis of PID control is proportional control; integral control can eliminate the steady-state error but may increase overshoot; derivative control can accelerate the response speed of a large inertia system and weaken the overshoot trend.

[0054] It should be noted that the chemical reactions involved in waterborne coatings are very sensitive to temperature. Too high or too low temperature may lead to unstable reaction rates, thus affecting the performance of the coatings. Too high temperature may accelerate the decomposition of the emulsion or evaporation of water in the waterborne coatings, resulting in unstable performance; while too low temperature may lead to incomplete reactions and the coatings cannot achieve the expected performance.

[0055] See Figure 2 , Figure 2 which is a schematic diagram of a waterborne coating production line provided by an embodiment of the present invention; there is a normal temperature range (such as 60°C - 70°C) for the chemical reactions during the production of waterborne coatings. Usually, by real-time monitoring the fluctuations of temperature data, when the temperature deviates from the normal temperature range required in the production stage, the PID controller will automatically adjust the heating or cooling equipment to ensure temperature stability. However, when the temperature is still within the normal temperature range required for the chemical reactions in the production stage, but the temperature change trend does not conform to the change law of the waterborne coating production reaction, it may indicate abnormal reaction rates or equipment failures. Therefore, this situation needs to be further analyzed to perform PID control in a timely manner.

[0056] The following will specifically describe the specific solution of an intelligent monitoring method for the waterborne coating production process based on data analysis provided by the present invention with reference to the accompanying drawings.

[0057] Please refer toFigure 1 , which shows a flowchart of an intelligent monitoring method for the production process of waterborne coatings based on data analysis provided by an embodiment of the present invention. The method includes:

[0058] S101: Periodically obtain the temperature data and concentration data of the materials in the reaction kettle at different sampling times; segment the time according to the extreme values of the temperature data to determine different temperature change cycle segments.

[0059] The scenario corresponding to the embodiment of the present invention is that the materials in the reaction kettle include the raw materials of waterborne coatings and reactants. The raw materials of waterborne coatings can specifically include acrylic acid and polyurethane, and the reactants can specifically be leveling agents, defoaming agents, wetting agents, thickening agents, etc. of waterborne coatings. The reactants are mixed with the raw materials of waterborne coatings and react in the reaction kettle to obtain waterborne coatings. It should be noted that the present invention aims to monitor and control the reaction stage in a timely manner, so that the obtained waterborne coatings are more stable.

[0060] It should be noted that in the production process of waterborne coatings, the chemical reactions involved in the preparation link are extremely sensitive to temperature. Temperature is the key factor controlling the reaction rate, affecting the quality consistency and performance of waterborne coatings. Only at an appropriate temperature can the chemical reaction proceed smoothly to ensure the quality of waterborne coatings. Temperature is closely related to the chemical reaction process. In the production process of waterborne coatings, as the main equipment, the reaction kettle provides a reaction space for the raw materials. The reaction process of waterborne coatings is a dynamic and continuous system with a clear periodicity. For example, in the initiation, chain growth, and termination stages of emulsion polymerization, its reaction rate and heat release characteristics will cause the temperature data to show specific periodic change characteristics.

[0061] Further explanation, the chemical reaction cycle in the production of waterborne coatings is as follows: In the initiation stage, a relatively high temperature (such as 60°C - 70°C) is required to activate the initiator and form free radicals; in the chain growth stage, the temperature is maintained stable to avoid abnormal reaction rate caused by temperature fluctuations; in the cooling stage, the temperature is gradually reduced to prevent the emulsion from becoming unstable due to excessive heat release.

[0062] Based on this, a high-precision temperature sensor PT100 platinum resistance thermometer can be selected to meet the temperature range (0°C - 120°C) of waterborne coating production. The sensor is installed at key production points, such as the inner wall of the reaction kettle and the outlet of the mixing tank, inside production containers, to ensure that the sensor is in close contact with the production environment and avoid measurement deviation.

[0063] Use a conductivity sensor to measure the concentration changes of specific reactants or products in the production process of waterborne coatings. Place the sensor at the inlet and outlet of the reaction kettle or in the circulation pipeline to ensure that the liquid flows through the detection area of the sensor.

[0064] To observe a rapidly changing chemical reaction, set the sampling frequency to collect the temperature value and concentration value of the materials in the reaction kettle once every 0.5 seconds, ensuring that the real-time dynamic changes of the reaction are captured. Transmit the collected values to the central control system, and perform data cleaning, noise reduction, time synchronization, and standardization processing on the collected values to unify the dimensions, obtaining temperature data and concentration data. Ensure the integrity, accuracy, and consistency of the temperature data and concentration data.

[0065] It should be noted that since the temperature data changes periodically, the period can be divided according to the extreme values.

[0066] Furthermore, in some embodiments of the present invention, time segments are divided according to the extreme values of the temperature data to determine different temperature change period segments, including: taking the time range between the minimum values of the two closest temperature data as a temperature change period segment.

[0067] See Figure 3 , Figure 3 which is a schematic diagram of the division of the temperature change period segment provided by an embodiment of the present invention. Among them, taking the minimum value as the cutting point, the time period between two adjacent minimum values is taken as a temperature change period segment. Then, the temperature value in the temperature change period segment fluctuates first increasing and then decreasing. The division of the temperature change period segment can effectively realize the periodic analysis of temperature fluctuations.

[0068] S102: Determine the temperature period characteristic index of each temperature change period segment according to the temperature fluctuations at each sampling moment in different temperature change period segments; determine the temperature data deviation index of each temperature change period segment according to the temperature fluctuations comparison and temperature period characteristic index difference between any temperature change period segment and all other temperature change periods.

[0069] In the embodiments of the present invention, the fluctuations of the temperature data conform to the periodic fluctuation characteristics, that is, under normal circumstances, the temperature changes caused by the release and absorption of heat during the chemical reaction process present a periodic waveform. Therefore, periodic analysis can be performed based on the temperature fluctuations.

[0070] Furthermore, in some embodiments of the present invention, the range of the temperature data at all sampling moments in the same temperature change period segment is taken as the first period index; the time interval from the starting moment of the temperature change period segment to any sampling moment is taken as the target interval corresponding to the sampling moment; calculate the ratio of the target interval to the duration of the temperature change period segment as the second period index corresponding to the sampling moment; determine the standard period coefficient of the temperature change period segment according to the second period index and the first period index at all sampling moments in the same temperature change period segment. The corresponding calculation formula is:

[0071] ; where, represents the standard period coefficient of the i-th temperature change period segment, represents the first period index of the i-th temperature change period segment, represents the second period index at the j-th sampling moment in the i-th temperature change period segment, represents pi, and K represents the number of all sampling moments in the i-th temperature change period segment.

[0072] Among them, the first period index represents the range, that is, the value of the first period index represents the fluctuation range, while the second period index represents the time proportion of the corresponding moment. Through the analysis of the sin function, the time proportion is substituted into the sine function and subtracted by to conform to the characteristic that the temperature change period segment starts from the minimum point; represents the temperature value at the j-th sampling moment under the standard sine fluctuation. Thus, according to the average value of the temperature values at all sampling moments, the standard period coefficient under the condition of conforming to the sine change is calculated.

[0073] Therefore, based on the difference between the actual temperature data and the standard period coefficient in the temperature change period segment, characteristic analysis is carried out. The absolute value of the difference between the average value of the temperature data at all sampling moments in the temperature change period segment and the standard period coefficient is calculated as the standard temperature difference; the opposite number of the standard temperature difference is subjected to maximum-minimum normalization processing to obtain the temperature period characteristic index of the temperature change period segment.

[0074] The larger the value of the standard temperature difference, the greater the difference between the actual situation and the standard situation of the sine function. The opposite number of it is normalized to obtain the temperature period characteristic index; the larger the value of the temperature period characteristic index, the more it conforms to the sine fluctuation characteristic.

[0075] It can be understood that since not all reactions strictly conform to the sine change characteristic and different reactions have certain differential effects, but in the same reaction kettle, the temperature changes of all temperature change period segments should be similar. Therefore, in order to analyze and obtain the temperature change period segment with a larger abnormality, the difference analysis of the temperature period characteristic indexes of all temperature change period segments is carried out, and the similarity analysis of the temperature fluctuations of different temperature change period segments is carried out, so as to obtain the temperature data deviation index of the temperature change period segment.

[0076] Further, in some embodiments of the present invention, according to the temperature fluctuation comparison and the difference in temperature cycle characteristic indexes between any temperature change cycle segment and all other temperature change cycles, the temperature data deviation index of each temperature change cycle segment is determined, including: calculating the DTW value of the temperature values within any two temperature change cycle segments based on the dynamic time warping algorithm, performing maximum-minimum normalization on the DTW value to obtain the first deviation coefficient of the temperature fluctuations within the two temperature change cycle segments; taking the absolute value of the difference between the temperature cycle characteristic indexes of any two temperature change cycle segments as the second deviation coefficient of the temperature fluctuations within the two temperature change cycle segments; calculating the product of the first deviation coefficient and the second deviation coefficient, and performing maximum-minimum normalization to obtain the cycle deviation index of the two temperature change cycle segments; taking the mean of the cycle deviation indexes of any temperature change cycle segment and all other temperature change cycle segments as the temperature data deviation index corresponding to any temperature change cycle segment.

[0077] Among them, the dynamic time warping algorithm is an algorithm well-known to those skilled in the art. Through the dynamic time warping algorithm, the fluctuation similarity analysis of two sequences in time series can be carried out. The DTW value is the numerical value obtained by the dynamic time warping algorithm. The larger the DTW value, the smaller the similarity between the two time series, which means that the two time series have a larger fluctuation deviation. Therefore, by performing maximum-minimum normalization on the DTW value, the first deviation coefficient of the temperature fluctuations within the two temperature change cycle segments is obtained. The larger the numerical value of the first deviation coefficient, the greater the difference in temperature fluctuations within the two temperature change cycle segments.

[0078] Among them, the second deviation coefficient represents the difference in temperature cycle characteristic indexes. The larger the numerical value of the second deviation coefficient, the greater the difference in cycle characteristics between the two temperature change cycle segments.

[0079] Therefore, by combining the first deviation coefficient and the second deviation coefficient, the cycle deviation index of the two temperature change cycle segments is obtained, and the mean of the cycle deviation indexes of any temperature change cycle segment and all other temperature change cycle segments is calculated as the temperature data deviation index corresponding to any temperature change cycle segment. That is, the larger the numerical value of the temperature data deviation index, the greater the difference in temperature fluctuations of the corresponding temperature change cycle segment compared with other temperature change cycle segments, that is, the greater the abnormality of the temperature change in this temperature change cycle segment.

[0080] S103: Determine the chemical reaction rate at each sampling moment according to the change of the concentration data at adjacent sampling moments; divide the sampling moments into two time categories, the pre-reaction period and the post-reaction period, according to the temperature data peak value, cluster the numerical values of the chemical reaction rates at different sampling moments in each time category to obtain time clusters; determine the chemical reaction deviation index of each sampling moment according to the difference in the chemical reaction rates between any sampling moment and other sampling moments in its time cluster.

[0081] In the embodiments of the present invention, since the concentration change is directly related to the chemical reaction rate, the chemical reaction rate can be directly determined according to the concentration change.

[0082] Further, in some embodiments of the present invention, determining the chemical reaction rate at each sampling moment according to the change in concentration data at adjacent sampling moments includes: taking any sampling moment as the target moment, and determining the two sampling moments closest to the target moment as the moments to be analyzed; taking the average value of the absolute values of the differences between the concentration data of the target moment and the concentration data of each moment to be analyzed as the chemical reaction rate at the target moment, and adjusting the target moment to obtain the chemical reaction rate at each sampling moment.

[0083] Among them, the greater the change in concentration between the target moment and the moment to be analyzed, the stronger the chemical reaction at the corresponding target moment, that is, the greater the chemical reaction rate.

[0084] Whether in the early stage or the late stage of the reaction, there are two stages, one is the rapid change stage and the other is the stable stage. That is to say, in the early stage of the reaction, first, the materials in the reaction kettle are in a stable stage due to high concentration but the temperature conditions are not met. Then, when gradually approaching the peak value of the temperature data, it means that the temperature is gradually rising, the reaction rate becomes faster, and the reaction gradually increases to the maximum value. At this time, it is in the rapid change stage; the same is true for the late stage of the reaction, which gradually degenerates from the rapid change stage to the stable stage. Therefore, in order to facilitate classification and analysis of different stages, clustering is required to achieve effective classification.

[0085] Further, in some embodiments of the present invention, clustering the numerical values of the chemical reaction rates at different sampling moments in each time category to obtain time clusters includes: setting the value of k to 2, and performing k-means clustering on the numerical values of the chemical reaction rates at all sampling moments in the same time category to obtain 2 clustering clusters in each time category as time clusters.

[0086] Among them, k-means clustering is a well-known clustering algorithm in the art. In the embodiments of the present invention, each time category is divided into two types: the rapid change stage and the stable stage. Therefore, the corresponding value of k can be set to 2 and divided into two types of time clusters. Two time clusters are obtained by clustering in the early stage of the reaction, and two time clusters are also obtained by clustering in the late stage of the reaction. That is, 4 different time clusters are obtained in the two time categories.

[0087] Then, specific analysis is performed according to the chemical reaction rates of the same time clusters.

[0088] Further, in some embodiments of the present invention, a chemical reaction deviation index for each sampling moment is determined according to the difference in chemical reaction rates between any sampling moment and other sampling moments in the time cluster to which it belongs, including: taking the mean value of the chemical reaction rates of all sampling moments in the time cluster as the first deviation judgment factor, and taking the median as the second deviation judgment factor; taking the absolute value of the difference between the chemical reaction rate of any sampling moment in the time cluster and the first deviation judgment factor of the time cluster as the first deviation coefficient; taking the absolute value of the difference between the chemical reaction rate of any sampling moment in the time cluster and the second deviation judgment factor of the time cluster as the second deviation coefficient; calculating the product of the first deviation coefficient and the second deviation coefficient, and performing maximum-minimum normalization processing as the chemical reaction deviation index of the sampling moment.

[0089] That is to say, for the chemical reaction rate at any sampling moment, the greater the difference between it and the median and the mean value, the greater the corresponding deviation index. The first deviation coefficient represents the difference between it and the mean value, and the second deviation coefficient represents the difference between it and the median value. Therefore, calculate the product of the first deviation coefficient and the second deviation coefficient, and perform maximum-minimum normalization processing as the chemical reaction deviation index of the sampling moment.

[0090] Among them, the chemical reaction deviation index represents the abnormality of the chemical reaction rate itself at the corresponding sampling moment. The larger the value of the chemical reaction deviation index, the higher the degree of abnormality of the chemical reaction rate at the corresponding sampling moment.

[0091] S104: Combine the temperature data deviation index and the chemical reaction deviation index to determine the temperature adjustment degree for each sampling moment; determine the necessity of temperature control for each sampling moment according to the change in the temperature adjustment degree within the preset neighborhood time sequence range for each sampling moment.

[0092] Since the temperature data deviation index represents the abnormality of the temperature change in the corresponding temperature change period segment, and the chemical reaction deviation index represents the abnormality of the chemical reaction rate itself at the corresponding sampling moment. The temperature data deviation index and the chemical reaction deviation index can be combined to analyze the overall abnormality at each sampling moment.

[0093] Further, in some embodiments of the present invention, combining the temperature data deviation index and the chemical reaction deviation index to determine the temperature adjustment degree for each sampling moment includes: calculating the product of the chemical reaction deviation index at the same sampling moment and the temperature data deviation index of the temperature change period segment in which the sampling moment is located to obtain the temperature adjustment degree corresponding to the sampling moment.

[0094] In the embodiments of the present invention, since the larger the value of the temperature data deviating from the index, the greater the temperature fluctuation difference of the corresponding temperature change period segment compared with other temperature change period segments, that is, the greater the abnormality of the temperature change in this temperature change period segment, and the larger the value of the chemical reaction deviating from the index, the higher the abnormal degree of the chemical reaction rate at the corresponding sampling moment.

[0095] Therefore, directly calculate the product of the chemical reaction deviation index and the temperature data deviation index to obtain the temperature adjustment degree at the corresponding sampling moment. The higher the temperature adjustment degree, the more abnormal the temperature change and the chemical reaction rate change at the corresponding sampling moment, and the more intervention is needed for regulation.

[0096] Further, in some embodiments of the present invention, according to the change of the temperature adjustment degree within the preset neighborhood time series range at each sampling moment, determine the necessity of temperature regulation at each sampling moment, including: taking any sampling moment as the target moment, and determining the two sampling moments before the target moment in the time series as the regulation influence moments; calculating the mean value of the temperature adjustment degree of the target moment and the temperature adjustment degrees of all regulation influence moments, and performing maximum-minimum normalization processing to obtain the temperature regulation necessity of the target moment.

[0097] It can be understood that the temperature regulation necessity is a numerical index representing the necessary degree of temperature regulation at the corresponding sampling moment. Due to a single abnormal change, it may be a special reason, while an abnormality lasting for multiple sampling moments can effectively represent the objectively existing abnormal effect in the time series. Therefore, in the embodiments of the present invention, the two sampling moments before the target moment in the time series are used as the regulation influence moments, so as to directly calculate the mean value of the temperature adjustment degree and perform maximum-minimum normalization processing to obtain the temperature regulation necessity, that is, the temperature regulation necessity takes into account the abnormalities of multiple consecutive sampling moments, avoids data acquisition errors, or instantaneous errors caused by special reasons, and improves the overall regulation effect.

[0098] S105: Perform temperature PID control adjustment in the reaction kettle according to the temperature regulation necessity at the current moment.

[0099] Since the temperature regulation necessity is a numerical index representing the necessary degree of temperature regulation at the corresponding sampling moment, therefore, control adjustment can be performed according to the temperature regulation necessity. The present invention uses a PID controller for control adjustment.

[0100] Further, in some embodiments of the present invention, performing temperature PID control adjustment in the reaction kettle according to the temperature regulation necessity at the current moment includes: when the temperature regulation necessity is greater than the preset necessity threshold, trigger the temperature PID control adjustment in the reaction kettle.

[0101] Among them, the preset necessity threshold is the threshold value for the necessity of temperature regulation. Optionally, the preset necessity threshold can specifically be 0.6. That is to say, when the necessity of temperature regulation is greater than 0.6, the PID controller can be controlled to start for control adjustment, so that the PID controller can promptly identify abnormalities, trigger the adaptive adjustment of the temperature by the PID controller, ensure that the chemical reaction is always in the best state, and avoid problems such as heat transfer lag or imbalance, and ensure the temperature consistency of the entire production process.

[0102] The process of the PID controller controlling the temperature in the waterborne coating production process is to dynamically adjust the control output according to the temperature deviation. At this time, the temperature deviation is replaced by the necessity of temperature regulation, so as to dynamically adjust the output of the heating or cooling equipment according to the control principles of proportional (P), integral (I), and derivative (D). That is to say, the necessity of temperature regulation is directly used as the start-stop condition for PID control adjustment, thus avoiding the problem of slow response of PID control adjustment.

[0103] The present invention periodically obtains temperature data and concentration data, and then, according to the numerical fluctuations of the temperature data in different temperature change cycle segments, determines the temperature data deviation index of each temperature change cycle segment. The temperature data deviation index characterizes the data abnormality of the temperature data in the corresponding temperature change cycle segment; then, determines the chemical reaction rate according to the change of the concentration data, and determines the chemical reaction deviation index of each sampling moment according to the chemical reaction rate distribution of each sampling moment in the same time cluster. The chemical reaction deviation index represents the abnormality of the chemical reaction at the corresponding sampling moment; in summary, combines the temperature data deviation index and the chemical reaction deviation index to determine the temperature adjustment degree, further determines the necessity of temperature regulation, and performs temperature PID control adjustment in the reaction kettle according to the necessity of temperature regulation. The present invention effectively analyzes the reaction state at each moment by combining the data information of multiple dimensions of temperature and concentration in the reaction kettle, avoids the problem of untimely overall regulation and slow response of PID control adjustment caused by only relying on temperature deviation, ensures that the chemical reaction is always in the best state, and guarantees the temperature consistency of the entire production process.

[0104] It should be noted that: the above-mentioned sequence of the embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or continuous order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0105] Each embodiment in this specification is described in a progressive manner. The same or similar parts between the embodiments can be referred to each other, and the key points of each embodiment are the differences from other embodiments.

Claims

1. An intelligent monitoring method for the production process of waterborne coatings based on data analysis, characterized in that, The method includes: Periodically obtaining the temperature data and concentration data of the materials in the reactor at different sampling times; segmenting time according to the extreme values of the temperature data to determine different temperature change cycle segments; Determining the temperature cycle characteristic index of each temperature change cycle segment according to the temperature fluctuations at each sampling time in different temperature change cycle segments; determining the temperature data deviation index of each temperature change cycle segment according to the temperature fluctuation comparison and temperature cycle characteristic index difference between any temperature change cycle segment and all other temperature change cycles; Determining the chemical reaction rate at each sampling time according to the change of the concentration data at adjacent sampling times; dividing the sampling times into two time categories, namely the pre-reaction period and the post-reaction period, according to the temperature data peak value, clustering the numerical values of the chemical reaction rates at different sampling times in each time category to obtain time clusters; determining the chemical reaction deviation index of each sampling time according to the chemical reaction rate difference between any sampling time and other sampling times in its time cluster; Combining the temperature data deviation index and the chemical reaction deviation index to determine the temperature adjustment degree at each sampling time; determining the necessity of temperature control at each sampling time according to the change of the temperature adjustment degree within the preset neighborhood time series range of each sampling time; Performing temperature PID control adjustment in the reactor according to the necessity of temperature control at the current time; Determining the temperature cycle characteristic index of each temperature change cycle segment according to the temperature fluctuations at each sampling time in different temperature change cycle segments, including: Taking the range of the temperature data of all sampling times in the same temperature change cycle segment as the first cycle index; Taking the time interval from the start time of the temperature change cycle segment to any sampling time as the target interval corresponding to the sampling time; calculating the ratio of the target interval to the duration of the temperature change cycle segment as the second cycle index corresponding to the sampling time; Determining the standard cycle coefficient of the temperature change cycle segment according to the second cycle index and the first cycle index of all sampling times in the same temperature change cycle segment, and the corresponding calculation formula is: ; where, represents the standard period coefficient of the i-th temperature change period segment, represents the first period index of the i-th temperature change period segment, represents the second period index at the j-th sampling moment in the i-th temperature change period segment, represents pi, and K represents the number of all sampling moments in the i-th temperature change period segment; Calculating the absolute value of the difference between the mean value of the temperature data of all sampling times in the temperature change cycle segment and the standard cycle coefficient as the standard temperature difference; Performing maximum-minimum normalization processing on the opposite number of the standard temperature difference to obtain the temperature cycle characteristic index of the temperature change cycle segment.

2. The intelligent monitoring method for the production process of water-based coatings based on data analysis according to claim 1, wherein, Segmenting time according to the extreme values of the temperature data to determine different temperature change cycle segments, including: Taking the time range between the two closest minimum values of the temperature data as a temperature change cycle segment.

3. The intelligent monitoring method for the production process of waterborne coatings based on data analysis according to claim 1, wherein, Determining the temperature data deviation index of each temperature change cycle segment according to the temperature fluctuation comparison and temperature cycle characteristic index difference between any temperature change cycle segment and all other temperature change cycles, including: Calculating the DTW value of the temperature values within any two temperature change cycle segments based on the dynamic time warping algorithm, and performing maximum-minimum normalization processing on the DTW value to obtain the first deviation coefficient of the temperature fluctuations within the two temperature change cycle segments; Take the absolute value of the difference between the temperature cycle characteristic indexes of any two temperature change cycle segments as the second deviation coefficient of the temperature fluctuation within the two temperature change cycle segments; Calculate the product of the first deviation coefficient and the second deviation coefficient, and perform maximum-minimum normalization to obtain the cycle deviation index of the two temperature change cycle segments; Take the mean of the cycle deviation indexes of any one temperature change cycle segment and all other temperature change cycle segments as the temperature data deviation index corresponding to any one temperature change cycle segment.

4. An intelligent monitoring method for the production process of water-based coatings based on data analysis according to claim 1, characterized in that, Determine the chemical reaction rate at each sampling moment according to the change of the concentration data at adjacent sampling moments, including: Take any sampling moment as the target moment, and determine the two sampling moments closest to the target moment as the moments to be analyzed; Take the mean of the absolute values of the differences between the concentration data at the target moment and the concentration data at each moment to be analyzed as the chemical reaction rate at the target moment, and adjust the target moment to obtain the chemical reaction rate at each sampling moment.

5. An intelligent monitoring method for the production process of waterborne coatings based on data analysis according to claim 1, characterized in that, Cluster the values of the chemical reaction rates at different sampling moments in each time category to obtain time clusters, including: Set the value of k to 2, and perform k-means clustering on the values of the chemical reaction rates at all sampling moments in the same time category to obtain 2 clustering clusters in each time category as the time clusters.

6. The intelligent monitoring method for the production process of water-based coatings based on data analysis according to claim 1, wherein, Determine the chemical reaction deviation index of each sampling moment according to the difference in the chemical reaction rates between any sampling moment and other sampling moments in the time cluster it belongs to, including: Take the mean of the chemical reaction rates at all sampling moments in the time cluster as the first deviation judgment factor, and take the median as the second deviation judgment factor; Take the absolute value of the difference between the chemical reaction rate at any sampling moment in the time cluster and the first deviation judgment factor of the time cluster as the first deviation coefficient; Take the absolute value of the difference between the chemical reaction rate at any sampling moment in the time cluster and the second deviation judgment factor of the time cluster as the second deviation coefficient; Calculate the product of the first deviation coefficient and the second deviation coefficient, and perform maximum-minimum normalization as the chemical reaction deviation index of the sampling moment.

7. The intelligent monitoring method for the production process of waterborne coatings based on data analysis according to claim 1, characterized in that, Combine the temperature data deviation index and the chemical reaction deviation index to determine the temperature adjustment degree at each sampling moment, including: Calculate the product of the chemical reaction deviation index at the same sampling moment and the temperature data deviation index of the temperature change cycle segment where the sampling moment is located to obtain the temperature adjustment degree corresponding to the sampling moment.

8. The intelligent monitoring method for the production process of waterborne coatings based on data analysis according to claim 1, characterized in that, Determine the necessity of temperature control at each sampling moment according to the change of the temperature adjustment degree within the preset neighborhood time series range of each sampling moment, including: Take any sampling moment as the target moment, and determine the two sampling moments before the target moment in the time series as the control influence moments; Calculate the mean of the temperature adjustment degrees at the target moment and all control influence moments, and perform maximum-minimum normalization to obtain the necessity of temperature control at the target moment.

9. The intelligent monitoring method for the production process of waterborne coatings based on data analysis according to claim 1, characterized in that, Perform temperature PID control adjustment in the reaction kettle according to the necessity of temperature control at the current moment, including: When the necessity of temperature control is greater than the preset necessity threshold, trigger the PID control adjustment of the temperature in the reaction kettle.

Images