Intelligent control method for cooling crystalline grain size

Abstract

The application relates to a cooling crystallization particle size intelligent control method, and relates to the technical field of crystallization particle size control.The method comprises the following steps: collecting the supersaturation of a solution in a target crystallization kettle and a crystal image sequence in real time, extracting current particle size distribution characteristics, and obtaining raw material impurity monitoring values and stirrer power fluctuation coefficients; a multi-section adaptive cooling curve is constructed; the target crystallization kettle is controlled to be cooled, and the collecting, constructing and controlling steps are iteratively executed until the crystallization process is completed, so that initial crystal slurry is obtained; the particle size distribution of the initial crystal slurry is measured, solid-liquid separation and drying treatment are performed, and finally, the crystal product is obtained.The application solves the problems that the traditional crystallization control relies on experience or a fixed mode, lacks dynamic adaptation capability, is difficult to cope with various disturbances, and leads to uneven product particle size and poor quality stability.

Description

Technical Field

[0001] This application relates to the field of crystallization particle size control, and in particular to a method for intelligent control of cooling crystallization particle size. Background Technology

[0002] As industrial manufacturing upgrades towards higher precision and higher quality, the uniformity of particle size and the stability of quality of crystal products have become core technical requirements, directly affecting the downstream application effect and product competitiveness, and placing higher demands on the ability to precisely control the crystallization process.

[0003] Currently, traditional crystallization control relies heavily on fixed parameter curves or human experience, employing an open-loop control mode. This makes it impossible to dynamically respond to various disturbances in production, resulting not only in large fluctuations in crystal particle size distribution but also in difficulty in accurately controlling product quality, thus failing to meet the stringent requirements of high-end manufacturing. Summary of the Invention

[0004] This application provides a method for intelligent control of cooling crystallization particle size, which improves the problem that traditional crystallization control relies on fixed curves or manual experience, making it difficult to cope with various disturbances, resulting in unstable crystal particle size distribution and difficulty in controlling product quality.

[0005] The embodiments of this application disclose the following technical solutions: This application provides a method for intelligent control of cooling crystallization particle size, the method comprising: Based on the compensated distribution peak position and half-width at half-maximum, and referring to the raw material impurity monitoring value and the stirrer power fluctuation coefficient, the phase transition activity coefficient is calculated; The difference between the real-time acquired oversaturation and the preset target oversaturation value is calculated as the real-time oversaturation deviation. Based on the phase transition activity coefficient, an adjustment factor is calculated, and the real-time supersaturation deviation is corrected to construct a cooling rate compensation function. Based on the preset crystallization temperature range, multiple adaptive cooling curves are divided and generated. The cooling rate of each cooling segment is determined by the calculation result of the cooling rate compensation function under the initial supersaturation condition.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application proposes an intelligent control method for cooling crystallization particle size. By real-time acquisition of the supersaturation of the solution and crystal image sequences within the target crystallizer, the current particle size distribution characteristics are extracted, and the raw material impurity monitoring values ​​and stirrer power fluctuation coefficient are simultaneously acquired. Multi-segment adaptive cooling curves are constructed, and the crystallizer temperature is controlled according to these curves through iterative optimization. Finally, qualified crystal products are obtained through particle size determination and closed-loop processing, achieving particle size control in the cooling crystallization process of high-purity potassium nitrate. First, relevant data is acquired in real-time, and particle size distribution characteristics, including the compensated distribution peak position and half-width at half-maximum (WHM), are extracted. Next, the phase transition activity coefficient is calculated based on raw material impurities and stirring parameters, a cooling rate compensation function is constructed, and multi-segment adaptive cooling curves are generated. Then, the temperature is controlled according to the curves, with iterative acquisition and curve reconstruction steps after each cooling segment. Initial crystal slurry is obtained after crystallization. The target particle size distribution range is then determined based on the crystallization process data, and the particle size of the initial crystal slurry is measured. Finally, qualified slurry undergoes solid-liquid separation and drying to obtain the final product, while unqualified slurry is returned to the dissolution process for recrystallization until the particle size meets the standard.

[0007] The technical solution of this application solves the problems of uneven particle size distribution and low product qualification rate caused by raw material fluctuations and changes in operating conditions during the traditional cooling crystallization process, and realizes intelligent control of the particle size of high-purity potassium nitrate cooling crystallization. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 A flowchart illustrating an intelligent control method for cooling crystallization particle size provided in an embodiment of this application; Figure 2 This is a schematic diagram of the process for constructing a multi-segment adaptive cooling curve provided in an embodiment of this application. Detailed Implementation

[0010] This application provides a method for intelligent control of cooling crystallization particle size, which solves the technical problem that the traditional crystallization process in the prior art relies on open-loop control with fixed cooling curves or human experience, and cannot cope with disturbances such as raw material fluctuations and equipment status changes, resulting in unstable crystal particle size distribution and difficulty in controlling product quality.

[0011] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0012] In the description of this application, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this application, "multiple" means two or more, unless otherwise explicitly specified.

[0013] In the description of this application, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this application is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. Details are set forth in the following description for purposes of explanation. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes will not be described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed in this application.

[0014] Examples, as shown in the appendix Figure 1 As shown, this application provides a method for intelligent control of cooling crystallization particle size, the method comprising the following steps: S110: Real-time acquisition of supersaturation and crystal image sequence of solution in target crystallizer, extraction of current particle size distribution characteristics, and simultaneous acquisition of raw material impurity monitoring values ​​and stirrer power fluctuation coefficient; In this embodiment of the application, during the crystallization process of producing high-purity potassium nitrate, the crystal particle size is dynamically affected by the solution state, raw material quality, and equipment operating status. In scenarios where there are differences in raw material composition fluctuations and stirring conditions between different batches of production, in order to capture the real-time status of the crystallization process, it is necessary to comprehensively collect core parameters, crystal morphology information, and related auxiliary monitoring data in the crystallization vessel to ensure the targeted nature of subsequent control.

[0015] Specifically, the supersaturation and crystal image sequence of the high-purity potassium nitrate production solution in the target crystallizer are first collected in real time. Then, through continuous frame difference analysis, grayscale morphology processing and data compensation calculation, the current particle size distribution characteristics that characterize the concentration trend and dispersion degree of crystal particles are extracted from the crystal image sequence to reflect the crystal growth state during the crystallization process.

[0016] The method provided in this application embodiment extracts the current granularity distribution features, including: The crystal image sequence is subjected to continuous frame difference analysis and grayscale morphological processing to segment and obtain the crystal contour set at each time step; Based on the set of crystal profiles, the equivalent circle diameter of each profile is calculated, and a real-time particle size distribution histogram is generated. Extract the distribution peak positions and the full width at half maximum (FWHM) from the real-time granularity distribution histogram; By combining the synchronously acquired oversaturation, the oversaturation deviation of the distribution peak position is compensated to obtain the compensated distribution peak position; The compensated distribution peak position and the distribution half-width together constitute the current particle size distribution characteristics.

[0017] Specifically, the crystal image sequence collected during the production of high-purity potassium nitrate is first processed to further segment and obtain a set of crystal outlines at each moment. The crystal image sequence is acquired by an industrial camera installed at the sight glass of the crystallization vessel, and is continuously collected at a preset sampling frequency to ensure dynamic recording of the growth morphology changes of potassium nitrate crystals.

[0018] Furthermore, continuous frame difference analysis is performed on the acquired image sequence. By comparing the pixel differences between adjacent frames, background noise and static interference are effectively filtered out. Then, grayscale morphological processing technology is used to enhance the edge clarity of the crystal contour, thereby segmenting the crystal contour set at each moment to ensure that the contour information can fully reflect the actual shape of the crystal.

[0019] Secondly, based on the set of crystal contours obtained from segmentation, the equivalent circle diameter of each contour is calculated. Specifically, the calculation of the equivalent circle diameter is based on the area of ​​the contour, using the formula... This parameter demonstrates that it can intuitively quantify the actual size of the crystal.

[0020] Furthermore, the equivalent circle diameter data of all crystal profiles are statistically integrated, and statistical frequencies are divided according to grain size intervals to form a real-time grain size distribution histogram. The horizontal axis of the histogram represents the equivalent circle diameter interval, and the vertical axis represents the proportion of crystals in the corresponding interval, to present the grain size distribution of the current batch of crystals.

[0021] Meanwhile, the distribution peak positions and half-width at half-maximum (WHM) of the distribution were extracted from the real-time particle size distribution histogram to characterize the concentration trend and dispersion degree of potassium nitrate crystal particles, respectively.

[0022] The distribution peak position corresponds to the equivalent circle diameter value with the highest frequency in the histogram, which can reflect the current concentration trend of crystal grain size. For example, when the distribution peak position is 0.8 mm, it indicates that the grain size of most potassium nitrate crystals is concentrated around 0.8 mm.

[0023] Furthermore, the full width at half maximum (FWHM) of the distribution peak is the width of the particle size interval corresponding to the half-peak values ​​on both sides of the distribution peak. It reflects the degree of dispersion of crystal particle size; the smaller the FWHM, the more uniform the particle size. For example, a FWHM of 0.15 mm indicates that the fluctuation range of crystal particle size near the concentration tendency is small.

[0024] Furthermore, it is necessary to combine the supersaturation data collected simultaneously to compensate for the deviation in the distribution peak position, correct the influence of supersaturation changes on the crystal growth rate, obtain a compensated distribution peak position that is closer to the actual growth state, and improve the accuracy of the particle size distribution characteristics.

[0025] The supersaturation is calculated by obtaining the real-time concentration of the solution using an online concentration sensor installed in the crystallization vessel and comparing it with pre-stored equilibrium concentration data of potassium nitrate at different temperatures. The calculation formula is "Supersaturation = Actual Concentration - Equilibrium Concentration". For example, if the equilibrium concentration of potassium nitrate at 28℃ is 40.1 g / 100 g water, and the real-time concentration is 42.3 g / 100 g water, then the supersaturation is 2.2 g / 100 g water.

[0026] Meanwhile, since supersaturation directly affects the crystal growth rate and may cause deviations in the distribution peak position, a compensation method of supersaturation correction coefficient is introduced by establishing a linear fitting relationship between supersaturation and grain size deviation. The distribution peak position is corrected according to the real-time supersaturation value to obtain the compensated distribution peak position, so as to ensure that the parameters can truly reflect the actual growth state of the crystal.

[0027] For example, the supersaturation correction factor for potassium nitrate crystallization was determined in advance through experiments to be 0.03 mm / (g·100g water). If the real-time supersaturation is 2.2 g / 100g water and the original distribution peak position is 0.75 mm, the calculated "compensated distribution peak position = 0.75 mm + 2.2 g / 100g water × 0.03 mm / (g·100g water)" finally yields a compensated distribution peak position of 0.82 mm, which is consistent with the actual concentrated particle size of the crystals.

[0028] Finally, the compensated distribution peak position and the distribution half-width at half-maximum together constitute the current particle size distribution characteristics. The two parameters analyze the crystal particle size state from the two dimensions of central tendency and dispersion, respectively, providing a data basis for the subsequent construction of adaptive cooling curves.

[0029] While extracting the current particle size distribution characteristics, the raw material impurity monitoring value and the stirrer power fluctuation coefficient are simultaneously obtained to comprehensively capture the key external factors affecting the potassium nitrate crystallization process.

[0030] In the method provided in this application embodiment, the raw material impurity monitoring value is the percentage concentration of key impurity ions in the current batch of feed, and the stirrer power fluctuation coefficient is the ratio of the standard deviation to the average value of the target crystallizer stirrer operating power data within a preset sampling period.

[0031] Specifically, the raw material impurity monitoring value is detected and output in real time by an online component analysis instrument connected to the raw material feed pipeline. This value is the percentage concentration of key impurity ions in the current batch of feed. In the production of high-purity potassium nitrate, key impurity ions include at least sodium ions.

[0032] For example, when the online component analyzer detects a sodium ion concentration percentage of 0.04%, this value is the raw material impurity monitoring value for the current batch, which can intuitively reflect the potential impact of raw material purity on crystal growth.

[0033] Meanwhile, the agitator power fluctuation coefficient is obtained by a power acquisition unit connected to the motor control system of the agitator in the target crystallizer. This unit records the real-time operating power of the agitator at a preset sampling period. In each sampling period, the standard deviation and average value of all power data are calculated, and the ratio of the two is the agitator power fluctuation coefficient.

[0034] For example, if the sampling period is set to 2 minutes, the average operating power of the stirrer recorded during this period is 15kW, and the standard deviation is 0.3kW, then the stirrer power fluctuation coefficient is 0.02. This coefficient can effectively reflect the stability of the stirring condition. If the coefficient is too large, it may indicate that there is crystal agglomeration or uneven local concentration in the crystallization vessel.

[0035] Meanwhile, during the acquisition of raw material impurity monitoring values ​​and agitator power fluctuation coefficients, if the raw material impurity monitoring values ​​exceed the preset normal range, it is necessary to check whether impurities have been mixed into the raw material feed pipeline and to re-collect and test the data. In addition, if the agitator power data is missing or shows abnormal sudden changes, it is necessary to supplement it by interpolation of data from adjacent periods or to repair the power acquisition unit and then re-collect the data to ensure the accuracy, completeness, and timeliness of both sets of data.

[0036] S120: Based on the current particle size distribution characteristics, raw material impurity monitoring values ​​and stirrer power fluctuation coefficient, construct multiple adaptive cooling curves, wherein each adaptive cooling curve is associated with a cooling rate compensation function. In this embodiment of the application, in order to generate a cooling strategy that can adapt to the real-time crystallization state and avoid the inability of fixed cooling mode to cope with the particle size instability caused by disturbance, it is necessary to combine multi-dimensional data to calculate key coefficients, construct compensation functions and divide cooling segments to ensure the accuracy and flexibility of the cooling process.

[0037] Specifically, based on the compensated distribution peak position and distribution half-width, combined with the raw material impurity monitoring value and the stirrer power fluctuation coefficient, the phase transformation activity coefficient is obtained by calculating the relative deviation ratio of each parameter, the comprehensive deviation product and the fourth root, so as to quantify the dynamic activity of the current crystallization process.

[0038] Furthermore, the difference between the real-time supersaturation and the preset target supersaturation value is calculated to obtain the real-time supersaturation deviation, providing a basis for adjusting the cooling rate.

[0039] Furthermore, an adjustment factor is calculated based on the phase change activity coefficient, where the phase change activity coefficient and the adjustment factor are positively correlated. The cooling rate adjustment value is then obtained by multiplying the real-time supersaturation deviation and the adjustment factor, and a cooling rate compensation function is constructed by combining the preset benchmark cooling rate.

[0040] Finally, based on the preset crystallization temperature range, the number of cooling sections and the initial sub-temperature range are determined by combining the phase transformation activity coefficient. The range boundary is corrected by the position of the compensated distribution peak and the raw material impurity monitoring value. The cooling rate compensation function is associated with each cooling section and the corresponding cooling rate is calculated. The cooling sections and cooling rates are summarized in order from high to low temperature to form a multi-segment adaptive cooling curve, which provides the execution basis for crystallization cooling control.

[0041] Step S120 in the method provided in this application embodiment includes: Based on the compensated distribution peak position and half-width at half-maximum, and referring to the raw material impurity monitoring value and the stirrer power fluctuation coefficient, the phase transition activity coefficient is calculated; The difference between the real-time acquired oversaturation and the preset target oversaturation value is calculated as the real-time oversaturation deviation. Based on the phase transition activity coefficient, an adjustment factor is calculated, and the real-time supersaturation deviation is corrected to construct a cooling rate compensation function. Based on the preset crystallization temperature range, multiple adaptive cooling curves are divided and generated. The cooling rate of each cooling segment is determined by the calculation result of the cooling rate compensation function under the initial supersaturation condition.

[0042] In this embodiment of the application, in order to adapt the cooling control strategy to the dynamic changes in the high-purity potassium nitrate crystallization process, it is necessary to calculate key coefficients through multi-dimensional parameters, construct compensation functions, and divide dynamic cooling sections to achieve fine control of the crystallization process and ensure the particle size uniformity and quality stability of the final product.

[0043] Specifically, the phase transformation activity coefficient is first calculated. This coefficient needs to be quantified by comprehensively considering factors such as the crystal growth state, raw material quality, and equipment operating conditions during the cooling crystallization process, in order to reflect the dynamic activity level of the current crystallization process.

[0044] The method provided in this application embodiment calculates the phase transition activity coefficient based on the compensated distribution peak position and half-width at half-maximum, and with reference to the raw material impurity monitoring value and the stirrer power fluctuation coefficient, including: The normal ranges of the compensated distribution peak position, distribution half width at half maximum, raw material impurity monitoring value, and agitator power fluctuation coefficient were obtained respectively. The relative deviation ratios of the four parameters relative to the center values ​​of their corresponding normal ranges are calculated respectively. The position of the compensated distribution peak corresponds to the first relative deviation ratio, the half-width at half-maximum of the distribution corresponds to the second relative deviation ratio, the raw material impurity monitoring value corresponds to the third relative deviation ratio, and the agitator power fluctuation coefficient corresponds to the fourth relative deviation ratio. Multiply the first relative deviation ratio, the second relative deviation ratio, the third relative deviation ratio, and the fourth relative deviation ratio to obtain the comprehensive deviation product; Calculate the fourth root of the comprehensive deviation product as the phase transition activity coefficient.

[0045] Specifically, the normal ranges of four parameters—the position of the distribution peak after compensation, the half-width at half-maximum (WHM) of the distribution, the monitoring value of raw material impurities, and the power fluctuation coefficient of the agitator—are first determined to establish a benchmark for judging the degree of parameter deviation, providing a basis for the subsequent calculation of the relative deviation ratio and the phase change activity coefficient.

[0046] For example, based on long-term data statistics and practical experience in the production process of high-purity potassium nitrate, the normal range of the peak position after compensation is 120μm-150μm, the normal range of the half-width at half-maximum is 15μm-25μm, the normal range of the raw material impurity monitoring value is 0.01%-0.05%, and the normal range of the stirrer power fluctuation coefficient is 0.01-0.03.

[0047] Further, the relative deviation ratio of each parameter is calculated. Specifically, the relative deviation ratio is calculated as "(actual value of parameter - center value of normal range) / center value of normal range". For example, if the actual value of the peak position after compensation is 138 μm, and the corresponding center value of the normal range is 135 μm, then the first relative deviation ratio is (138-135) / 135≈0.022. If the actual value of the full width at half maximum (FWHM) of the distribution is 22 μm, and the corresponding center value of the normal range is 20 μm, then the second relative deviation ratio is (22-20) / 20=0.1.

[0048] Furthermore, if the actual raw material impurity monitoring value is 0.03%, and the corresponding normal range center value is 0.03%, then the third relative deviation ratio is 0. If the actual agitator power fluctuation coefficient is 0.025, and the corresponding normal range center value is 0.02, then the fourth relative deviation ratio is (0.025-0.02) / 0.02=0.25.

[0049] Furthermore, the four relative deviation ratios are multiplied together to obtain the comprehensive deviation product. For example, in the above example, the comprehensive deviation product = 0.022 × 0.1 × 0 × 0.25 = 0. The fourth root of this comprehensive deviation product is then calculated, and the phase transformation activity coefficient is finally obtained as 0, indicating that the crystallization process is stable and there is no obvious disturbance.

[0050] Conversely, if the product of the overall deviations is 0.000125, its fourth root is approximately 0.186, and the corresponding phase transformation activity coefficient is 0.186. This indicates that the crystallization process is slightly disturbed by factors such as raw material impurities and stirring conditions, and the subsequent cooling rate needs to be adjusted to adapt to this dynamic change.

[0051] Furthermore, the real-time supersaturation deviation is calculated. By comparing the real-time supersaturation value with the preset target supersaturation value, the difference between the two is determined as the real-time supersaturation deviation. This deviation directly reflects the degree of deviation between the current solution crystallization driving force and the ideal state, providing basic data support for the precise adjustment of the cooling rate. For example, if the preset target supersaturation value is 2.5 g / 100 g water, and the real-time supersaturation value is 3.2 g / 100 g water, then the real-time supersaturation deviation is 0.7 g / 100 g water.

[0052] Furthermore, an adjustment factor is calculated based on the phase transition activity coefficient, and a weighted correction is applied to the real-time supersaturation deviation to construct a cooling rate compensation function. This function dynamically adapts to the changes in the active state of the high-purity potassium nitrate crystallization process, ensuring that the cooling rate accurately responds to the supersaturation deviation.

[0053] The method provided in this application embodiment calculates an adjustment factor based on the phase transition activity coefficient, corrects the real-time supersaturation deviation, and constructs a cooling rate compensation function, including: Based on the phase transition activity coefficient, an adjustment factor is calculated, wherein the larger the phase transition activity coefficient, the larger the adjustment factor is obtained; Multiply the real-time oversaturation deviation by the adjustment factor to obtain the cooling rate adjustment value; The preset baseline cooling rate is added to the cooling rate adjustment value, and the sum is the output value of the cooling rate compensation function.

[0054] First, the adjustment factor is calculated. This adjustment factor is derived based on the phase transition activity coefficient and the preset gain coefficient. Its core function is to transform the dynamic active state of the crystallization process into an adjustment factor for the cooling rate, ensuring that the adjustment range of the cooling rate matches the degree of process disturbance.

[0055] Specifically, the formula for calculating the adjustment factor is "adjustment factor = 1 + gain coefficient × phase change activity coefficient", where the gain coefficient is used to adjust the response sensitivity of the control process to changes in phase change activity. It is determined through statistical analysis of historical data from the production of high-purity potassium nitrate and can be set to a value of 0.5.

[0056] For example, if the calculated phase transformation activity coefficient is 0.2 and the gain coefficient is 0.5, then the adjustment factor = 1 + 0.5 × 0.2 = 1.1, indicating that there is a slight disturbance in the current crystallization process, and the cooling rate needs to be slightly adjusted. If the phase transformation activity coefficient is 0.6, the adjustment factor = 1 + 0.5 × 0.6 = 1.3, indicating that the degree of process disturbance is large, and the adjustment range of the cooling rate needs to be increased to quickly adapt to the change in crystallization state.

[0057] Furthermore, the cooling rate adjustment value is calculated by multiplying the real-time supersaturation deviation by an adjustment factor. The real-time supersaturation deviation reflects the degree of deviation between the current solution crystallization driving force and the ideal state, while the adjustment factor quantifies the need for cooling rate adjustment due to the disturbance. The product of the two can accurately determine the specific amount of cooling rate adjustment.

[0058] For example, if the preset target supersaturation value is 2.8 g / 100 g water and the real-time supersaturation is 3.5 g / 100 g water, then the real-time supersaturation deviation = 3.5 - 2.8 = 0.7 g / 100 g water; if the adjustment factor is 1.2, then the cooling rate adjustment value = 0.7 × 1.2 = 0.84 °C / h, which is the cooling rate adjustment range adapted to the current crystallization state.

[0059] Finally, a cooling rate compensation function is constructed, which adds the preset baseline cooling rate to the adjusted cooling rate value. The final sum is the output value of the cooling rate compensation function. The preset baseline cooling rate is derived from long-term statistical analysis and practical experience of high-purity potassium nitrate production process data, and represents the basic cooling rate under stable conditions during the crystallization process. For example, the baseline cooling rate is set to 2.2℃ / h.

[0060] Based on the example above, the output value of the cooling rate compensation function is 2.2 + 0.84 = 3.04℃ / h, which is the cooling rate adapted to the current state. If the real-time supersaturation is lower than the preset target supersaturation value, the real-time supersaturation deviation is negative, and the cooling rate adjustment value is also negative. The final output cooling rate will be lower than the reference cooling rate to slow down the cooling rate and avoid uneven particle size caused by slow crystal growth.

[0061] Furthermore, after the cooling rate compensation function is constructed, the temperature range is divided according to the preset crystallization temperature range, and multiple adaptive cooling curves are generated to dynamically adapt to the real-time changes of the crystallization system and ensure uniform and stable crystal particle size.

[0062] As attached Figure 2 As shown, the method provided in this application embodiment divides and generates multiple adaptive cooling curves according to a preset crystallization temperature range, including: Obtain the reference temperature span and, in conjunction with the phase transition activity coefficient, calculate the unit control temperature span. Calculate the ratio of the preset crystallization temperature range to the unit control temperature span, and round it up to get the number of cooling sections. The crystallization temperature range is divided into consecutive initial sub-temperature ranges, with the same number of sub-temperature ranges as the cooling sections. Based on the compensated distribution peak position and the raw material impurity monitoring value, the boundary temperature of the continuous initial sub-temperature range is dynamically corrected to generate a continuous corrected sub-temperature range; For each cooling segment corresponding to a continuous correction sub-temperature range, the cooling rate compensation function is associated, and the cooling rate of the current cooling segment is calculated and determined based on the real-time oversaturation at the start of the cooling segment. According to the continuous correction sub-temperature range from high temperature to low temperature, the cooling segments and the calculated cooling rates are summarized in sequence to form multiple adaptive cooling curves.

[0063] First, calculate the unit control temperature span. This unit control temperature span needs to be derived by combining the reference temperature span and the phase transformation activity coefficient. Its purpose is to rationally divide the cooling control temperature units based on the dynamic activity of the crystallization process.

[0064] Specifically, the formula for calculating the unit control temperature span is "Unit control temperature span = Reference temperature span / Phase change activity coefficient". The reference temperature span is obtained through long-term statistical analysis of high-purity potassium nitrate production process data and practical experience, for example, 10℃.

[0065] For example, if the calculated phase transformation activity coefficient is 0.8 and the reference temperature span is 10℃, then the unit control temperature span = 10 / 0.8 = 12.5℃, indicating that the current crystallization process has a moderate level of activity, and a larger temperature span can be used as the control range for a single cooling section. If the phase transformation activity coefficient is 1.2, the unit control temperature span = 10 / 1.2 ≈ 8.33℃, indicating a high level of activity, and the temperature span of a single cooling section needs to be reduced to improve the response sensitivity of the cooling control.

[0066] Furthermore, the number of cooling sections is calculated by ratioing the preset crystallization temperature range length to the unit control temperature span and rounding the result up. The preset crystallization temperature range also needs to be determined in conjunction with the crystallization process requirements of high-purity potassium nitrate, for example, 60℃-30℃, with a range length of 30℃.

[0067] Based on the above examples, if the unit control temperature span is 12.5℃, then the number of cooling sections = 30 / 12.5 = 2.4, which is rounded up to 3 sections; if the unit control temperature span is 8.33℃, then the number of cooling sections = 30 / 8.33 ≈ 3.6, which is rounded up to 4 sections, to ensure that the division of cooling sections can completely cover the entire crystallization temperature range.

[0068] Furthermore, the crystallization temperature range is divided into consecutive initial sub-temperature ranges, the same number as the number of cooling sections. For example, if the preset crystallization temperature range is 60℃-30℃ and there are 3 cooling sections, it is divided into three consecutive initial sub-temperature ranges: 60℃-50℃, 50℃-40℃, and 40℃-30℃. If there are 4 cooling sections, it is divided into four consecutive initial sub-temperature ranges: 60℃-52.5℃, 52.5℃-45℃, 45℃-37.5℃, and 37.5℃-30℃, laying the foundation for subsequent temperature boundary correction and cooling rate allocation.

[0069] Furthermore, based on the compensated distribution peak position and raw material impurity monitoring value, the boundary temperature of the continuous initial sub-temperature range is dynamically corrected to generate a continuous corrected sub-temperature range, so as to ensure that the cooling control is more in line with the actual crystallization state.

[0070] The method provided in this application embodiment dynamically corrects the boundary temperature of the continuous initial sub-temperature range based on the compensated distribution peak position and the raw material impurity monitoring value, generating a continuously corrected sub-temperature range, including: Obtain the expected growth temperature offset corresponding to the position of the compensated distribution peak, wherein the larger the position of the compensated distribution peak, the larger the positive value of the corresponding expected growth temperature offset. The phase change temperature compensation amount corresponding to the raw material impurity monitoring value is obtained, wherein the higher the raw material impurity monitoring value, the larger the positive value of the corresponding phase change temperature compensation amount. The desired growth temperature offset is added to the phase transition temperature compensation to obtain the total temperature correction. Using the total temperature correction amount, the boundary temperature values ​​of each of the continuous initial sub-temperature intervals are shifted and adjusted in the same direction to form a continuous corrected sub-temperature interval.

[0071] First, the desired growth temperature offset is calculated. This desired growth temperature offset is derived based on the difference between the compensated distribution peak position and the target grain size distribution peak position, so that the temperature range adapts to the actual growth trend of the current crystal.

[0072] Specifically, the formula for calculating the desired growth temperature offset is "Desired growth temperature offset = k × (Compensated distribution peak position - Target particle size distribution peak position)". Here, k is the temperature-size sensitivity coefficient (positive value), determined through statistical analysis of historical process data from high-purity potassium nitrate production, and can be taken as 0.02℃ / μm. Furthermore, the target particle size distribution peak position is also obtained through statistical analysis and experience using high-purity potassium nitrate production process data, for example, 135μm.

[0073] For example, if the compensated distribution peak position is 142 μm and the target particle size distribution peak position is 135 μm, with k = 0.02℃ / μm, then the expected growth temperature offset is 0.02 × (142 - 135) = 0.14℃, indicating that the current crystal growth size is slightly larger than the target value, and the temperature range needs to be adjusted to adapt to this growth trend. If the compensated distribution peak position is 130 μm and the target particle size distribution peak position is 135 μm, then the expected growth temperature offset is 0.02 × (130 - 135) = -0.1℃, indicating that the crystal growth size is slightly smaller than the target value, and the temperature range needs to be adjusted in the opposite direction to promote reasonable crystal growth.

[0074] Furthermore, the phase transition temperature compensation amount is calculated. This compensation amount needs to be calculated based on the raw material impurity monitoring value and the preset compensation coefficient to compensate for the influence of impurities in the raw material on the crystallization phase transition temperature.

[0075] Specifically, the formula for calculating the phase change temperature compensation is "Phase change temperature compensation = p × raw material impurity monitoring value". Where p is the impurity-temperature compensation coefficient (positive value), determined through experimental data and process experience, for example, 0.3℃ / %.

[0076] For example, if the raw material impurity monitoring value is 0.04% and p = 0.3℃ / %, then the phase transformation temperature compensation amount is 0.3 × 0.04 = 0.012℃. If the raw material impurity monitoring value is 0.06%, then the phase transformation temperature compensation amount is 0.3 × 0.06 = 0.018℃. The higher the impurity content, the greater the compensation amount, in order to offset the interference of impurities on the phase transformation temperature and ensure the stability of the crystallization process.

[0077] Furthermore, the desired growth temperature offset is added to the phase transition temperature compensation to obtain the total temperature correction. For example, in the above example, the desired growth temperature offset is 0.14℃, and the phase transition temperature compensation is 0.012℃, then the total temperature correction = 0.14 + 0.012 = 0.152℃. If the desired growth temperature offset is -0.1℃, and the phase transition temperature compensation is 0.018℃, then the total temperature correction = -0.1 + 0.018 = -0.082℃. The total temperature correction integrates the influence of both crystal growth trend and raw material impurities, providing a basis for temperature range adjustment.

[0078] Finally, the total temperature correction is used to shift the boundary temperature value of each boundary temperature range in the same direction. That is, all boundary temperature values ​​are increased or decreased synchronously by the total temperature correction to ensure that the temperature range as a whole adapts to the crystal growth trend of high-purity potassium nitrate and the influence of raw material impurities, so that the temperature range of each cooling section is more in line with the actual crystallization state.

[0079] For example, if the preset continuous initial sub-temperature ranges are 60℃-50℃, 50℃-40℃, and 40℃-30℃, and the total temperature correction is 0.152℃, then the adjusted continuous corrected sub-temperature ranges are 60.152℃-50.152℃, 50.152℃-40.152℃, and 40.152℃-30.152℃, respectively. By adjusting in the same direction, the temperature range of each sub-temperature range is ensured to be adapted to the current crystallization state, matching the crystal growth trend and offsetting the adverse effects of raw material impurities, making the subsequent application of cooling rates more targeted.

[0080] Furthermore, after the continuous correction sub-temperature range is generated, a cooling rate compensation function is associated with the cooling segment corresponding to each continuous correction sub-temperature range, and the cooling rate of the current cooling segment is calculated and determined based on the real-time supersaturation at the start of the cooling segment, so as to ensure that the rate of each cooling segment can be adapted to the crystallization state of the corresponding temperature range.

[0081] Specifically, for each cooling section corresponding to a continuous correction sub-temperature range, when the cooling section is about to start, the supersaturation data of the solution in the target crystallizer is collected in real time. The real-time supersaturation data is input into the cooling rate compensation function, and the cooling rate of the current cooling section is obtained through function calculation.

[0082] For example, the first continuous correction temperature range is 60.152℃-50.152℃, and the real-time supersaturation collected at the beginning of the corresponding cooling section is 3.2g / 100g water. After calculation by the cooling rate compensation function, the cooling rate of this cooling section is 2.8℃ / h.

[0083] Furthermore, the second continuous correction temperature range is 50.152℃-40.152℃, with an initial real-time supersaturation of 2.7 g / 100 g water, resulting in a calculated cooling rate of 2.3℃ / h. The third continuous correction temperature range is 40.152℃-30.152℃, with an initial real-time supersaturation of 2.1 g / 100 g water, resulting in a calculated cooling rate of 1.7℃ / h.

[0084] Furthermore, following the order of continuous correction sub-temperature ranges from high to low, each cooling segment and the calculated cooling rate are sequentially summarized to form a multi-segment adaptive cooling curve. For example, the above three cooling segments are integrated in the order of 60.152℃-50.152℃ (2.8℃ / h), 50.152℃-40.152℃ (2.3℃ / h), and 40.152℃-30.152℃ (1.7℃ / h) to form a complete multi-segment adaptive cooling curve.

[0085] Ultimately, the obtained multi-segment adaptive cooling curves can dynamically adjust the cooling rate according to the real-time status of different temperature ranges during the crystallization process, effectively coping with disturbances caused by factors such as raw material impurities and stirring conditions, so as to ensure that high-purity potassium nitrate crystals are in a suitable cooling environment throughout the crystallization process, and ultimately achieve the goal of uniform and stable particle size distribution.

[0086] S130: The target crystallizer is cooled by using the adaptive cooling curve, and the acquisition, construction and control steps are iteratively executed after each cooling stage until the crystallization process is completed and the initial crystal slurry is obtained. In this embodiment of the application, in order to ensure that the cooling control of the entire crystallization process always meets the crystal growth requirements and avoid uneven particle size distribution caused by state drift, it is necessary to implement cooling step by step based on the constructed adaptive cooling curve, and dynamically adjust the control strategy through iterative optimization to ensure the particle size stability and consistency of the final crystal product.

[0087] Specifically, the cooling control process of the target crystallizer is first initiated. The constructed multi-segment adaptive cooling curve is imported into the temperature control system of the crystallizer. The system executes the cooling operation sequentially from high temperature to low temperature in each cooling segment. Each cooling segment uses the corresponding cooling rate as the control benchmark. By adjusting the heat exchange system parameters of the crystallizer, the cooling rate is controlled to ensure that the solution temperature in the cooling segment decreases smoothly according to the preset curve.

[0088] For example, for the preset temperature range of 60℃-30℃ for high-purity potassium nitrate crystals, if the adaptive cooling curve includes three cooling segments, namely 60.152℃-50.152℃ (2.8℃ / h), 50.152℃-40.152℃ (2.3℃ / h), and 40.152℃-30.152℃ (1.7℃ / h), the control system first reduces the temperature from 60.152℃ to 50.152℃ at a rate of 2.8℃ / h, thus completing the cooling task of the first cooling segment.

[0089] After each cooling stage, the current cooling operation is paused, and the data acquisition and curve reconstruction process is initiated. This involves real-time acquisition of supersaturation data and crystal image sequences of the solution inside the target crystallizer, while simultaneously obtaining the raw material impurity monitoring values ​​and agitator power fluctuation coefficients for that stage.

[0090] For example, after the first cooling stage, the real-time supersaturation was 2.7 g / 100 g water. New particle size distribution characteristics were extracted through image analysis. The raw material impurity monitoring value remained unchanged at 0.04%, and the stirrer power fluctuation coefficient was 0.023.

[0091] Furthermore, based on the newly acquired real-time data, the complete process of extracting granular distribution features, calculating phase transformation activity coefficients, constructing cooling rate compensation functions, and generating multi-segment adaptive cooling curves is repeated to form a new cooling curve that adapts to the current crystallization state, thereby realizing the dynamic updating of the control strategy.

[0092] Subsequently, based on the newly constructed adaptive cooling curve, the cooling control of the next cooling segment is initiated. For example, in the new cooling curve constructed based on the updated parameters, the temperature range of the second cooling segment is adjusted to 50.152℃-40.08℃, and the cooling rate is optimized to 2.4℃ / h. The control system continues to perform cooling according to these parameters to ensure that the cooling rate matches the current crystal growth state.

[0093] Throughout the crystallization process, the iterative process of "cooling down the cooling section - data acquisition - curve reconstruction - cooling down the next cooling section" is repeated until the temperature of the solution in the target crystallizer drops to the lower limit of the preset crystallization temperature range, at which point the crystallization process terminates. At this point, an initial crystal slurry containing high-purity potassium nitrate crystals is formed in the target crystallizer.

[0094] S140: The particle size distribution of the initial crystal slurry is measured. If the measured particle size distribution falls within the target particle size distribution range, the current batch crystallization process is determined to be complete. The initial crystal slurry is then subjected to solid-liquid separation and drying treatment to obtain the final crystal product.

[0095] In this embodiment of the application, in order to ensure that the final product meets the quality standards, it is necessary to verify the crystallization effect through particle size measurement, perform subsequent processing on qualified slurry, and implement cyclic optimization on unqualified slurry to ensure the consistency and stability of product particle size.

[0096] Before conducting particle size distribution determination, it is necessary to first determine the target particle size distribution range applicable to the current batch in order to establish a quality judgment standard that fits the current production conditions of high-purity potassium nitrate, and avoid misjudgment caused by the inability of a fixed range to adapt to batch differences.

[0097] In the method provided in this application embodiment, the step of determining the target particle size distribution range includes: During the crystallization process, the real-time supersaturation and stirrer power fluctuation coefficient of each cooling section are recorded simultaneously, and the average supersaturation and average stirrer power fluctuation coefficient of all cooling sections are calculated. Based on the average supersaturation, the stirrer power fluctuation coefficient, and the raw material impurity monitoring value, the adjustment amounts for the upper and lower limits of the basic particle size distribution range are calculated respectively. The adjustment amounts are superimposed on the upper and lower limits of the basic particle size distribution range to generate a target particle size distribution range suitable for the current batch.

[0098] Specifically, during the execution of each cooling section, supersaturation data of the solution is collected every 15 minutes by a supersaturation sensor deployed in the target crystallizer to ensure that the data can fully reflect the dynamic changes in the crystallization driving force within the cooling section.

[0099] Meanwhile, the power monitoring module of the agitator continuously collects operating power data and calculates the ratio of the standard deviation to the average value of the power data in each cooling section, which is used as the agitator power fluctuation coefficient for that cooling section to reflect the stability of the agitation operation.

[0100] After the crystallization process is complete, the supersaturation data of all cooling sections are collected, and the arithmetic mean is calculated as the average supersaturation. The agitator power fluctuation coefficients of all cooling sections are then weighted and averaged to obtain the average agitator power fluctuation coefficient.

[0101] Furthermore, the collected average supersaturation, average agitator power fluctuation coefficient, and raw material impurity monitoring values ​​were normalized. Normalization involves converting parameters of different dimensions and magnitudes into values ​​between 0 and 1 to eliminate the influence of differences in parameter units.

[0102] For example, the average supersaturation value range is determined based on historical production data as 1.0g / 100g water to 4.0g / 100g water. If the average supersaturation of the current batch is 2.5g / 100g water, then the normalized average supersaturation = (2.5-1.0) / (4.0-1.0) = 0.5. The historical range of the average agitator power fluctuation coefficient is 0.01-0.05. If the average agitator power fluctuation coefficient of the current batch is 0.03, then the normalized average agitator power fluctuation coefficient = (0.03-0.01) / (0.05-0.01) = 0.5. The historical range of the raw material impurity monitoring value is 0.01%-0.05%. If the raw material impurity monitoring value of the current batch is 0.03%, then the normalized raw material impurity monitoring value = (0.03%-0.01%) / (0.05%-0.01%) = 0.5.

[0103] Further, the adjustment amount is calculated based on the normalized parameters. Specifically, the formula for calculating the adjustment amount is: "Upper limit of adjustment amount = Baseline adjustment amount × (1 + Normalized average supersaturation) × (1 + Normalized average agitator power fluctuation coefficient) × (1 + Normalized raw material impurity monitoring value)" and "Lower limit of adjustment amount = -Baseline adjustment amount × (1 + Normalized average supersaturation) × (1 + Normalized average agitator power fluctuation coefficient) × (1 + Normalized raw material impurity monitoring value)". The baseline adjustment amount is determined to be 5.0 μm based on a combination of historical production data and quality inspection results.

[0104] For example, combining the normalized parameter values ​​above, the upper limit of the adjustment is 5.0×(1+0.5)×(1+0.5)×(1+0.5)=16.875μm, and the lower limit of the adjustment is -16.875μm.

[0105] Finally, the calculated adjustment amount is superimposed on the basic particle size distribution range to generate the target particle size distribution range for the current batch. The basic particle size distribution range is a preset fixed range, also determined based on the product application requirements and production process standards of high-purity potassium nitrate, for example, 120μm-150μm.

[0106] Based on the adjustments in the example above, the upper limit of the target particle size distribution range for the current batch is 150 + 16.875 = 166.875 μm, and the lower limit is 120 - 16.875 = 103.125 μm, meaning the target particle size distribution range is 103.125 μm - 166.875 μm. This range fully incorporates the dynamic characteristics of the current batch's production, ensuring consistent product quality while flexibly adapting to reasonable fluctuations during the production process, providing a scientific standard for determining the particle size of the subsequent initial crystal slurry.

[0107] Furthermore, after determining the target particle size distribution range, the initial crystal slurry is subjected to particle size distribution measurement, and the crystal particle size is verified to meet the quality standards of the current batch using professional testing equipment. Specifically, a laser particle size analyzer is used as the measuring device. This device needs to be calibrated in advance with standard particle samples to ensure that the measurement accuracy meets the quality testing requirements for high-purity potassium nitrate.

[0108] First, three initial crystal slurry samples were selected from different locations within the target crystallization vessel. Each sample was thoroughly stirred before being slowly injected into the analyzer's test cell to avoid crystal particle sedimentation affecting the test results. The equipment uses the principle of laser scattering to analyze the scattering angle and intensity of the laser light by the crystal particles, converting this into corresponding particle size distribution data. It outputs key parameters such as D10, D50, and D90, as well as a particle size distribution curve for each sample. The average value of the three sample test results was taken as the final particle size distribution determination result to reduce errors caused by single sampling or equipment fluctuations.

[0109] Furthermore, the final particle size distribution measurement results are compared with the determined target particle size distribution range. If the measured D50 value falls within the target range and the ratio of D10, D90 values ​​to D50 meets the preset requirements, then the crystallization process of the current batch is determined to be complete.

[0110] Furthermore, the qualified initial crystal slurry undergoes solid-liquid separation treatment using a centrifuge. The centrifugal force generated by high-speed rotation efficiently separates the crystal particles from the mother liquor. The separated wet crystals are then conveyed into a potassium dryer. At a temperature of 120℃-150℃, the dryer removes adsorbed moisture and residual mother liquor from the crystal surface. The dried crystals are then cooled to room temperature in a potassium cooler, ultimately yielding a high-purity potassium nitrate crystal product with uniform particle size.

[0111] The method provided in this application embodiment includes measuring the particle size distribution of the initial crystal slurry. If the measured particle size distribution falls within the target particle size distribution range, the current batch crystallization process is determined to be complete. The initial crystal slurry is then subjected to solid-liquid separation and drying to obtain the final crystal product. The method further includes: If the measured particle size distribution does not fall within the target particle size distribution range, the initial crystal slurry is returned to the dissolution process for re-dissolution to obtain a recycled raw material solution. Based on the raw material impurity monitoring value and stirrer power fluctuation coefficient of the regenerated raw material solution, the real-time acquisition of supersaturation and crystal image sequence, extraction of current particle size distribution characteristics, construction of multi-segment adaptive cooling curves and control of cooling process are re-executed to obtain new crystal slurry; The reprocessing process is iteratively executed until the particle size distribution measurement result of the new crystal slurry falls within the target particle size distribution range, at which point the loop stops; The new crystal slurry that meets the target particle size distribution range requirements is subjected to solid-liquid separation and drying to output the final crystal product.

[0112] If the measured particle size distribution does not fall within the target particle size distribution range, for example, if the D50 value is 98 μm, which is lower than the lower limit of the target range of 103.125 μm, it indicates that the current crystal particle size does not meet the requirements and rework is required. The crystal growth state is adjusted through closed-loop optimization of dissolution and recrystallization to ensure that the particle size of the final product meets the quality standards of high-purity potassium nitrate.

[0113] Specifically, the initial crystal slurry is returned to the potassium dissolving tank via a conveying pipeline, and an appropriate amount of deionized water is injected. Under heating and stirring conditions, the crystal particles are completely dissolved to obtain a regenerated raw material solution. For the regenerated raw material solution, its raw material impurity monitoring values ​​are first collected, and the concentration percentage of key impurity ions is determined using an online component analysis instrument.

[0114] Simultaneously, the operating power data of the agitator was recorded, and the power fluctuation coefficient of the agitator was calculated. Based on the obtained data, the steps of real-time acquisition of supersaturation and crystal image sequences were repeated. Data was collected every 15 minutes using a supersaturation sensor, and crystal image sequences were acquired and processed using a high-definition camera to extract new particle size distribution characteristics.

[0115] Furthermore, based on the new particle size distribution characteristics, the raw material impurity monitoring value of the regenerated raw material solution and the stirrer power fluctuation coefficient, the phase transformation activity coefficient is recalculated, the cooling rate compensation function and multi-segment adaptive cooling curve are constructed, and the target crystallization vessel is cooled according to the new cooling curve to generate new crystal slurry.

[0116] Furthermore, the particle size distribution of the new crystal slurry is measured again. If the measurement result still does not meet the standard, the closed-loop process of "returning to dissolution-recrystallization-particle size measurement" is repeated until the particle size distribution measurement result of the new crystal slurry falls within the target particle size distribution range, and the cycle is stopped.

[0117] For example, after the first rework process, the D50 value of the new crystal slurry is 105μm, which falls within the target range of 103.125μm-166.875μm, and the D10 and D90 values ​​meet the requirements. At this time, the new crystal slurry is processed according to the above solid-liquid separation and drying process to output the final crystal product.

[0118] Meanwhile, throughout the entire closed-loop process, detailed records of the process parameters for each rework are required, including dissolution temperature, stirring rate, and cooling curve parameters, to provide data support for subsequent optimization of the initial crystallization process and gradually reduce the defect rate.

[0119] Ultimately, through the above-mentioned testing and closed-loop rework optimization, the final crystal product obtained has a uniform particle size distribution and meets quality requirements, which can meet the needs of multiple fields for high-purity potassium nitrate, thereby improving the stability of the production process and the market competitiveness of the product.

[0120] The embodiments of this application, through the above specific implementation methods, achieve the following technical effects: This application proposes an intelligent control method for cooling crystallization particle size. First, the supersaturation of the solution in the target crystallizer and the crystal image sequence are collected in real time to extract the current particle size distribution characteristics, and the raw material impurity monitoring value and the stirrer power fluctuation coefficient are acquired simultaneously. Then, the phase transformation activity coefficient is calculated based on multi-dimensional data, a cooling rate compensation function is constructed, and multiple adaptive cooling curves are divided and generated. Then, the cooling curve is used to control the cooling of the crystallizer. After each cooling segment, the collection, construction and control steps are iteratively executed until crystallization is completed and the initial crystal slurry is obtained. Then, the target particle size distribution range is determined by combining the crystallization process data, and the particle size of the initial crystal slurry is measured. Finally, the qualified slurry is subjected to solid-liquid separation and drying to obtain the final product, and the unqualified slurry is returned to the dissolution process for recrystallization. The process is iteratively repeated until the particle size meets the standard.

[0121] The method provided in this application, through the technical solution of "real-time acquisition and feature extraction of multi-dimensional crystallization state data - calculation of phase transformation activity coefficient and construction of cooling rate compensation function - generation of multi-segment adaptive cooling curve and iterative cooling control - determination of dynamic target particle size range and closed-loop management of product quality", solves the problems of uneven particle size distribution and low product qualification rate caused by raw material fluctuations and changes in working conditions in the traditional cooling crystallization process. It realizes intelligent control of the particle size of high-purity potassium nitrate cooling crystallization, and ensures the stability and consistency of product quality.

[0122] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0123] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0124] This specification and accompanying drawings are merely illustrative examples of this application and are intended to cover any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application intends to include such modifications and variations.

Claims

1. A method for intelligent control of cooling crystallization particle size, characterized in that, The method includes: Real-time acquisition of supersaturation and crystal image sequences of the solution in the target crystallizer, extraction of current particle size distribution characteristics, and simultaneous acquisition of raw material impurity monitoring values ​​and stirrer power fluctuation coefficient; Based on the current particle size distribution characteristics, raw material impurity monitoring values ​​and stirrer power fluctuation coefficient, multiple adaptive cooling curves are constructed, wherein each adaptive cooling curve is associated with a cooling rate compensation function. The target crystallizer is cooled using the adaptive cooling curve, and the acquisition, construction and control steps are iteratively executed after each cooling stage until the crystallization process is completed to obtain the initial crystal slurry. The initial crystal slurry is subjected to particle size distribution measurement. If the measured particle size distribution falls within the target particle size distribution range, the current batch crystallization process is determined to be complete. The initial crystal slurry is then subjected to solid-liquid separation and drying treatment to obtain the final crystal product.

2. The intelligent control method for cooling crystallization particle size according to claim 1, characterized in that, Extract the current granularity distribution features, including: The crystal image sequence is subjected to continuous frame difference analysis and grayscale morphological processing to segment and obtain the crystal contour set at each time step; Based on the set of crystal profiles, the equivalent circle diameter of each profile is calculated, and a real-time particle size distribution histogram is generated. Extract the distribution peak positions and the full width at half maximum (FWHM) from the real-time granularity distribution histogram; By combining the synchronously acquired oversaturation, the oversaturation deviation of the distribution peak position is compensated to obtain the compensated distribution peak position; The compensated distribution peak position and the distribution half-width together constitute the current particle size distribution characteristics.

3. The intelligent control method for cooling crystallization particle size according to claim 1, characterized in that, The raw material impurity monitoring value is the percentage concentration of key impurity ions in the current batch of feed, and the stirrer power fluctuation coefficient is the ratio of the standard deviation to the average value of the target crystallizer stirrer operating power data within a preset sampling period.

4. The intelligent control method for cooling crystallization particle size according to claim 1, characterized in that, Based on the current particle size distribution characteristics, raw material impurity monitoring values, and agitator power fluctuation coefficient, a multi-segment adaptive cooling curve is constructed, including: Based on the compensated distribution peak position and half-width at half-maximum, and referring to the raw material impurity monitoring value and the stirrer power fluctuation coefficient, the phase transition activity coefficient is calculated; The difference between the real-time acquired oversaturation and the preset target oversaturation value is calculated as the real-time oversaturation deviation. Based on the phase transition activity coefficient, an adjustment factor is calculated, and the real-time supersaturation deviation is corrected to construct a cooling rate compensation function. Based on the preset crystallization temperature range, multiple adaptive cooling curves are divided and generated. The cooling rate of each cooling segment is determined by the calculation result of the cooling rate compensation function under the initial supersaturation condition.

5. The intelligent control method for cooling crystallization particle size according to claim 4, characterized in that, Based on the compensated peak position and half-width at half-maximum (FWHM), and referring to the raw material impurity monitoring value and the stirrer power fluctuation coefficient, the phase transition activity coefficient is calculated, including: The normal ranges of the compensated distribution peak position, distribution half width at half maximum, raw material impurity monitoring value, and agitator power fluctuation coefficient were obtained respectively. The relative deviation ratios of the four parameters relative to the center values ​​of their corresponding normal ranges are calculated respectively. The position of the compensated distribution peak corresponds to the first relative deviation ratio, the half-width at half-maximum of the distribution corresponds to the second relative deviation ratio, the raw material impurity monitoring value corresponds to the third relative deviation ratio, and the agitator power fluctuation coefficient corresponds to the fourth relative deviation ratio. Multiply the first relative deviation ratio, the second relative deviation ratio, the third relative deviation ratio, and the fourth relative deviation ratio to obtain the comprehensive deviation product; Calculate the fourth root of the comprehensive deviation product as the phase transition activity coefficient.

6. The intelligent control method for cooling crystallization particle size according to claim 4, characterized in that, Based on the phase transition activity coefficient, an adjustment factor is calculated, and the real-time supersaturation deviation is corrected. A cooling rate compensation function is constructed, including: Based on the phase transition activity coefficient, an adjustment factor is calculated, wherein the larger the phase transition activity coefficient, the larger the adjustment factor is obtained; Multiply the real-time oversaturation deviation by the adjustment factor to obtain the cooling rate adjustment value; The preset baseline cooling rate is added to the cooling rate adjustment value, and the sum is the output value of the cooling rate compensation function.

7. The intelligent control method for cooling crystallization particle size according to claim 4, characterized in that, Based on the preset crystallization temperature range, multiple adaptive cooling curves are divided and generated, including: Obtain the reference temperature span and, in conjunction with the phase transition activity coefficient, calculate the unit control temperature span. Calculate the ratio of the preset crystallization temperature range to the unit control temperature span, and round it up to get the number of cooling sections. The crystallization temperature range is divided into consecutive initial sub-temperature ranges, with the same number of sub-temperature ranges as the cooling sections. Based on the compensated distribution peak position and the raw material impurity monitoring value, the boundary temperature of the continuous initial sub-temperature range is dynamically corrected to generate a continuous corrected sub-temperature range; For each cooling segment corresponding to a continuous correction sub-temperature range, the cooling rate compensation function is associated, and the cooling rate of the current cooling segment is calculated and determined based on the real-time oversaturation at the start of the cooling segment. According to the continuous correction sub-temperature range from high temperature to low temperature, the cooling segments and the calculated cooling rates are summarized in sequence to form multiple adaptive cooling curves.

8. The intelligent control method for cooling crystallization particle size according to claim 7, characterized in that, Based on the compensated distribution peak position and the raw material impurity monitoring value, the boundary temperature of the continuous initial sub-temperature range is dynamically corrected to generate a continuously corrected sub-temperature range, including: Obtain the expected growth temperature offset corresponding to the position of the compensated distribution peak, wherein the larger the position of the compensated distribution peak, the larger the positive value of the corresponding expected growth temperature offset. The phase change temperature compensation amount corresponding to the raw material impurity monitoring value is obtained, wherein the higher the raw material impurity monitoring value, the larger the positive value of the corresponding phase change temperature compensation amount. The desired growth temperature offset is added to the phase transition temperature compensation to obtain the total temperature correction. Using the total temperature correction amount, the boundary temperature values ​​of each of the continuous initial sub-temperature intervals are shifted and adjusted in the same direction to form a continuous corrected sub-temperature interval.

9. The intelligent control method for cooling crystallization particle size according to claim 1, characterized in that, The steps for determining the target particle size distribution range include: During the crystallization process, the real-time supersaturation and stirrer power fluctuation coefficient of each cooling section are recorded simultaneously, and the average supersaturation and average stirrer power fluctuation coefficient of all cooling sections are calculated. Based on the average supersaturation, the stirrer power fluctuation coefficient, and the raw material impurity monitoring value, the adjustment amounts for the upper and lower limits of the basic particle size distribution range are calculated respectively. The adjustment amounts are superimposed on the upper and lower limits of the basic particle size distribution range to generate a target particle size distribution range suitable for the current batch.

10. The intelligent control method for cooling crystallization particle size according to claim 1, characterized in that, The initial crystal slurry is subjected to particle size distribution measurement. If the measured particle size distribution falls within the target particle size distribution range, the current batch crystallization process is considered complete. The initial crystal slurry is then subjected to solid-liquid separation and drying to obtain the final crystal product. The process also includes: If the measured particle size distribution does not fall within the target particle size distribution range, the initial crystal slurry is returned to the dissolution process for re-dissolution to obtain a recycled raw material solution. Based on the raw material impurity monitoring value and stirrer power fluctuation coefficient of the regenerated raw material solution, the real-time acquisition of supersaturation and crystal image sequence, extraction of current particle size distribution characteristics, construction of multi-segment adaptive cooling curves and control of cooling process are re-executed to obtain new crystal slurry; The reprocessing process is iteratively executed until the particle size distribution measurement result of the new crystal slurry falls within the target particle size distribution range, at which point the loop stops; The new crystal slurry that meets the target particle size distribution range requirements is subjected to solid-liquid separation and drying to output the final crystal product.

Images