Lithium hexafluorophosphate production control system and method

Abstract

The invention discloses a lithium hexafluorophosphate production control system and method thereof, and relates to the technical field of lithium hexafluorophosphate production, and specifically comprises the following steps: obtaining the crystallization dynamic information of each area inside the reactor during the lithium hexafluorophosphate crystallization process in real time, and analyzing it after obtaining, respectively generating the crystal growth driving force coefficient and the particle uniform distribution index of each area; constructing a particle size distribution evaluation model for the generated crystal growth driving force coefficient and the particle uniform distribution index of each area, generating the particle size distribution coefficient of each area, and performing a comprehensive analysis after the generation to generate an evaluation coefficient; analyzing the generated evaluation coefficient, and evaluating the uniformity of the particle size distribution inside the reactor during the lithium hexafluorophosphate crystallization process. The invention solves the problem of uneven particle size distribution during the lithium hexafluorophosphate crystallization process, and achieves the effect of improving the particle size uniformity and product quality stability.

Description

Technical Field

[0001] The invention relates to the technical field of lithium hexafluorophosphate production, and in particular to a lithium hexafluorophosphate production control system and a method thereof. Background Art

[0002] Lithium hexafluorophosphate is an important lithium salt electrolyte widely used in lithium-ion batteries. It has the advantages of good conductivity and high thermal stability, and is one of the key determinants of battery performance. However, its production process involves a variety of chemical reactions, including the reaction of phosphoric acid and lithium fluoride, the treatment of fluorine gas and the separation of by-products. These process links have extremely strict requirements on process parameters such as temperature, pressure, and reactant ratio. If the production process is not properly controlled, it may lead to reduced product purity, excessive residual by-products, and even cause safety problems, thereby affecting the electrochemical performance and service life of lithium batteries. Therefore, strict control of the production of lithium hexafluorophosphate can not only improve the stability and purity of product quality, but also optimize process efficiency, reduce defective product rate and production costs, meet the stringent requirements of high-performance lithium batteries for electrolyte materials, and provide a solid guarantee for the development of the new energy industry.

[0003] The existing lithium hexafluorophosphate production control technology mainly relies on advanced automated control systems and multi-parameter monitoring methods to ensure the stability of the production process and product quality. On the one hand, through online sensors and control modules, key process parameters such as reaction temperature, pressure, reactant concentration and reaction time are monitored in real time to ensure that the process conditions strictly meet the set range, thereby avoiding the generation of byproducts and reduced reaction efficiency caused by overheating, overpressure or imbalance of raw material ratios; on the other hand, an automated data acquisition and feedback adjustment system is used to analyze various data in the production process in real time. If a trend deviating from the process requirements is found, the operating parameters can be adjusted in time, such as adjusting the raw material flow or reaction environment, to restore to the optimal reaction conditions. In addition, in the product separation and purification stage, the impurity content is reduced through precise filtration and solvent evaporation control to ensure that the purity of lithium hexafluorophosphate meets battery-grade standards. These technologies work together to significantly improve the production consistency and quality stability of lithium hexafluorophosphate, while reducing energy consumption and defective product rates, laying the foundation for large-scale industrial production.

[0004] The prior art has the following deficiencies:

[0005] During the crystallization process of lithium hexafluorophosphate, when the temperature, concentration and fluid flow rate at different positions in the reactor are unevenly distributed, the particle growth rate in the local area will be significantly different. Since the existing technology usually relies only on single-point data (such as the temperature or concentration at the center of the reactor) to evaluate the overall particle size distribution, it ignores the dynamic changes at other positions and cannot reflect the actual impact of local conditions on particle growth. When the temperature in the local area is high or the flow rate is low, the particles may grow rapidly and form too large particles; while in areas with low temperature or too high flow rate, the particle growth rate may be too slow and form too small particles. This local loss of control is caused by the complexity of the fluid flow in the reactor and the unevenness of the reaction dynamics, especially when the stirring speed or fluid viscosity fluctuates slightly. The existing technology lacks a dynamic evaluation mechanism based on multi-point real-time data, and cannot identify the risk of local particle size distribution loss of control and adjust the stirring parameters or process conditions in time, resulting in uneven particle size distribution throughout the reactor. This will directly affect the product quality of lithium hexafluorophosphate, including reduced particle purity and insufficient particle size stability, making it difficult to meet the requirements of lithium batteries for uniform conductivity; at the same time, the subsequent separation and purification process will become more complicated, increasing production costs and further reducing production efficiency and quality control stability.

[0006] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0007] The object of the present invention is to provide a lithium hexafluorophosphate production control system and method thereof to solve the problems in the above-mentioned background technology.

[0008] In order to achieve the above object, the present invention provides the following technical solution: a lithium hexafluorophosphate production control method, specifically comprising the following steps:

[0009] During the crystallization process of lithium hexafluorophosphate, the interior of the reactor used for lithium hexafluorophosphate crystallization is evenly divided into several areas, and the temperature, concentration and fluid flow rate distribution uniformity of each area inside the reactor are detected in real time. When uneven conditions are detected in some areas, the particle size distribution uniformity analysis process is triggered;

[0010] Real-time acquisition of crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate, and analysis after acquisition to generate the crystal growth driving force coefficient and particle uniformity distribution index of each area;

[0011] A particle size distribution evaluation model is constructed based on the crystal growth driving force coefficient and particle uniformity distribution index of each generated region, and the particle size distribution coefficient of each region is generated. After the generation, a comprehensive analysis is performed to generate an evaluation coefficient;

[0012] The generated evaluation coefficients are analyzed to evaluate the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate, and the uniformity of the particle size distribution is divided into high uniformity, normal uniformity and non-uniformity according to the evaluation results, and a high uniformity signal, a normal uniformity signal and a non-uniform signal are generated respectively;

[0013] In the case of generating uneven signals, continuously obtain several evaluation coefficients generated subsequently, conduct comprehensive analysis, generate different levels of early warning signals of particle size distribution out of control risk, and take corresponding control measures for different early warning levels;

[0014] The crystallization dynamic information and particle size distribution coefficient of each area inside the reactor are continuously monitored, and the parameters and control strategies of the particle size distribution evaluation model are dynamically adjusted according to the real-time monitoring results. At the same time, historical data are stored and analyzed to optimize the model parameters and control strategies.

[0015] Preferably, the crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate is obtained in real time, and analyzed after acquisition to generate the crystal growth driving force coefficient and particle uniform distribution index of each area respectively, which specifically includes the following steps:

[0016] Real-time acquisition of crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate, and pre-processing after acquisition;

[0017] Extracting crystal growth dynamics information and particle distribution uniformity information from the pre-processed crystallization dynamics information of each region;

[0018] The extracted crystal growth dynamics information and particle distribution uniformity information are analyzed to generate the crystal growth driving force coefficient and particle uniformity distribution index of each region respectively.

[0019] Preferably, the logic for obtaining the crystal growth driving force coefficient of each region is as follows:

[0020] The crystal growth dynamics information in the crystallization dynamics information of each pre-processed region is extracted, including the temperature gradient, solution saturation and fluid shear stress change rate of each region inside the reactor at different times during the crystallization process of lithium hexafluorophosphate, and calibrated as , and , represents the temperature gradient of the i-th region inside the reactor at time m during the crystallization of lithium hexafluorophosphate. It represents the solution saturation of the i-th region inside the reactor at time m during the crystallization process of lithium hexafluorophosphate. represents the rate of change of fluid shear stress in the i-th region of the reactor at time m during a period of time during the crystallization of lithium hexafluorophosphate, i=1, 2, 3, ..., k, m=1, 2, 3, ..., g, k and g are both positive integers;

[0021] Calculate the crystal growth driving force coefficient of each region. The specific calculation formula is as follows:

[0022] ,

[0023] In the formula, CDC i is the crystal growth driving force coefficient of the ith region.

[0024] Preferably, the logic for obtaining the particle uniform distribution index of each region is as follows:

[0025] The particle distribution uniformity information in the crystallization dynamic information of each pre-processed area is extracted, including the concentration variance, flow velocity gradient and temperature fluctuation amplitude of each area inside the reactor at different times during the crystallization process of lithium hexafluorophosphate, and calibrated as , and , represents the concentration variance of the i-th region inside the reactor at time n during the crystallization process of lithium hexafluorophosphate. represents the flow velocity gradient of the i-th region inside the reactor at time n during the crystallization process of lithium hexafluorophosphate, represents the temperature fluctuation amplitude of the i-th region inside the reactor at time n during a period of time during the crystallization process of lithium hexafluorophosphate, i=1, 2, 3, ..., k, n=1, 2, 3, ..., h, k and h are both positive integers;

[0026] Calculate the particle uniform distribution index of each area. The specific calculation formula is as follows:

[0027] ,

[0028] Where, PDI i is the particle uniform distribution index of the ith region.

[0029] Preferably, the crystal growth driving force coefficient CDC for each region generated is i and particle distribution index PDI i Construct a particle size distribution evaluation model and generate the particle size distribution coefficient of each area by weighted summation according to the formula:

[0030] ,

[0031] In the formula, GDC iis the particle size distribution coefficient of the i-th region, ω1 and ω2 are the crystal growth driving force coefficients CDC of each region respectively i and particle distribution index PDI i The non-zero weight coefficient of , and ω1+ω2=1;

[0032] The particle size distribution coefficient GDC of each area i The evaluation coefficient ENC is generated by calculating its standard deviation according to the formula: .

[0033] Preferably, a preset evaluation coefficient threshold interval [ENC min , ENC max ], and after determination, it is compared with the generated evaluation coefficient ENC, and the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate is evaluated according to the comparison results, and the uniformity of the particle size distribution is divided into high uniformity, normal uniformity and non-uniformity according to the evaluation results, and a high uniformity signal, a normal uniformity signal and a non-uniform signal are generated respectively. The specific comparison analysis is as follows:

[0034] If ENC<ENC min , during the crystallization of lithium hexafluorophosphate, the uniformity of the particle size distribution inside the reactor is high, generating a high uniformity signal;

[0035] If ENC min ≤ENC≤ENC max , during the crystallization of lithium hexafluorophosphate, the uniformity of the particle size distribution inside the reactor is normal, and a normal uniform signal is generated;

[0036] If ENC>ENC max ,During the crystallization process of lithium hexafluorophosphate, the uniformity of the particle size distribution inside the reactor is uneven, generating an uneven signal.

[0037] Preferably, in the case of generating an uneven signal, several evaluation coefficients generated subsequently are continuously obtained and calibrated as ENC x , x represents the numbers of several evaluation coefficients subsequently generated when an uneven signal is generated, x=1, 2, 3, ..., d, d is a positive integer;

[0038] Calculate the subsequent generated evaluation coefficients ENC x The average ENC — , according to the formula: ;

[0039] Calculate the subsequent generated evaluation coefficients ENC x Standard deviation of ENC σ , according to the formula: ;

[0040] Determine a preset mean value threshold of a number of evaluation coefficients subsequently generated in the case of generating an inhomogeneous signal and preset standard deviation threshold , the average value of several evaluation coefficients ENC — and standard deviation ENC σ The preset average value thresholds of several evaluation coefficients are and preset standard deviation threshold Compare and generate different levels of early warning signals for the risk of particle size distribution out of control according to the comparison results, and take corresponding control measures for different early warning levels. The specific comparison and analysis are as follows:

[0041] like and , generating a high-level early warning signal for the risk of out-of-control particle size distribution. The control measures taken are: optimizing the fluid flow state by adjusting the stirring speed and the position of the stirring paddle to reduce the flow rate fluctuation, reducing the temperature gradient inside the reactor to reduce the unevenness of the thermal field distribution, and improving the uniformity of the solute distribution by adjusting the solution feed rate and concentration distribution. Real-time monitoring of the temperature, concentration and flow rate changes after control, and dynamically adjusting the control parameters according to the feedback results until the particle size distribution returns to a uniform state;

[0042] like and , generating an early warning signal for the risk of loss of control of the medium-level particle size distribution. The control measures taken are: gradually adjusting the temperature distribution of each area inside the reactor to reduce temperature fluctuations, balancing the flow rate difference by optimizing the stirring speed and changing the fluid direction, alleviating the non-uniformity of the solute concentration by compensation, and tracking the parameter changes of each area after control in real time. According to the monitoring results, the control strategy is optimized to ensure that the uniformity of the particle size distribution inside the reactor is gradually improved;

[0043] like and , generating a warning signal for the risk of low-level particle size distribution being out of control, and taking the following control measures: reducing the saturation of the overall solution in the reactor to reduce the crystal growth rate, reducing the fluid shear stress by slowing down the overall stirring speed to avoid growth disturbances, and reducing the impact of temperature changes on particle size distribution by adjusting the temperature distribution control range. At the same time, continuously monitoring the state inside the reactor, recording changes in particle size distribution, and ensuring that the overall uniformity is gradually improved;

[0044] like and , generate a warning signal of no risk of loss of control and do not take any regulatory measures.

[0045] Preferably, the lithium hexafluorophosphate production control system includes a region division and uniformity detection module, a crystallization dynamic information acquisition and parameter generation module, a particle size distribution evaluation and coefficient generation module, a uniformity evaluation and signal generation module, a loss of control risk warning and regulation module, and a real-time monitoring and model optimization module;

[0046] The regional division and uniformity detection module divides the interior of the reactor used for lithium hexafluorophosphate crystallization into several regions during the crystallization process of lithium hexafluorophosphate, and performs real-time detection of the temperature, concentration and fluid flow rate distribution uniformity of each region inside the reactor. When non-uniformity is detected in certain regions, the particle size distribution uniformity analysis process is triggered;

[0047] The crystallization dynamic information acquisition and parameter generation module acquires the crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate in real time, analyzes it after acquisition, and generates the crystal growth driving force coefficient and particle uniform distribution index of each area respectively;

[0048] The particle size distribution evaluation and coefficient generation module builds a particle size distribution evaluation model based on the crystal growth driving force coefficient and particle uniform distribution index of each region, generates the particle size distribution coefficient of each region, and performs a comprehensive analysis after the generation to generate the evaluation coefficient;

[0049] The uniformity evaluation and signal generation module analyzes the generated evaluation coefficients, evaluates the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate, and divides the uniformity of the particle size distribution into high uniformity, normal uniformity and non-uniformity according to the evaluation results, and generates high uniformity signals, normal uniformity signals and non-uniform signals respectively;

[0050] The out-of-control risk warning and control module, when generating uneven signals, continuously obtains several evaluation coefficients generated subsequently, conducts comprehensive analysis, generates different levels of out-of-control risk warning signals for particle size distribution, and takes corresponding control measures for different warning levels;

[0051] The real-time monitoring and model optimization module continuously monitors the crystallization dynamic information and particle size distribution coefficient of each area inside the reactor, and dynamically adjusts the parameters and control strategies of the particle size distribution evaluation model according to the real-time monitoring results. At the same time, it stores and analyzes historical data to optimize model parameters and control strategies.

[0052] The beneficial effects of the present invention are:

[0053] 1. The present invention realizes precise control of particle size distribution during the crystallization process of lithium hexafluorophosphate, significantly improving product quality and process efficiency. Specifically, by dividing the interior of the reactor into several areas and monitoring the distribution uniformity of temperature, concentration and flow rate in real time, it is possible to fully capture the key data of each area under dynamic conditions, solving the problem of particle size distribution evaluation distortion caused by relying only on single-point data in the prior art. In addition, based on the generated crystal growth driving force coefficient and particle uniformity distribution index, a particle size distribution evaluation model is constructed, and the evaluation coefficient is generated. The overall uniformity of the particle size distribution can be quantified with high precision, providing a scientific basis for subsequent regulation. These measures jointly ensure the uniformity and stability of the particle size of the lithium hexafluorophosphate product, thereby meeting the strict requirements of high-performance lithium batteries for raw material quality.

[0054] 2. The present invention solves the impact of differences in conditions in various areas inside the reactor on the particle size distribution through multi-point real-time data collection and dynamic analysis, thereby avoiding the phenomenon of local particle size out of control. By introducing partition monitoring and calculation models, the limitations of insufficient overall evaluation accuracy in traditional methods are overcome, and efficient control of the uniformity of particle size distribution is achieved. Secondly, by setting up a dynamic evaluation model and early warning mechanisms of different levels, control measures can be triggered in real time according to the evaluation results, and graded responses can be taken according to the risk level, greatly improving the adaptability and reliability of the system. This hierarchical control method not only avoids the waste of resources caused by excessive adjustment, but also prevents quality problems caused by insufficient control, and optimizes production costs and process efficiency.

[0055] 3. The present invention adopts a modular architecture in design, including a regional division and uniformity detection module, a crystallization dynamic information acquisition and parameter generation module, a particle size distribution evaluation and coefficient generation module, etc. This modular design is not only easy to implement, but also has good scalability, providing flexible adjustment space for subsequent process optimization. At the same time, the storage and analysis of historical data and the dynamic optimization function of the evaluation model in the present invention can provide personalized control strategies for the production conditions of different batches, thereby improving the stability and repeatability of the production process. In general, the present invention not only fundamentally solves the key problems in the prior art, but also significantly enhances the quality control capability and industrial application value of lithium hexafluorophosphate production through a full-process closed-loop control mechanism and efficient data utilization. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 The figure is a schematic flow chart of the lithium hexafluorophosphate production control method of the present invention.

[0057] Figure 2 This is a module schematic diagram of the lithium hexafluorophosphate production control system of the present invention. DETAILED DESCRIPTION

[0058] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0059] The present invention provides Figure 1 The lithium hexafluorophosphate production control method shown specifically comprises the following steps:

[0060] During the crystallization process of lithium hexafluorophosphate, the interior of the reactor used for lithium hexafluorophosphate crystallization is evenly divided into several areas, and the temperature, concentration and fluid flow rate distribution uniformity of each area inside the reactor are detected in real time. When uneven conditions are detected in some areas, the particle size distribution uniformity analysis process is triggered;

[0061] During the crystallization process of lithium hexafluorophosphate, in order to more accurately monitor the dynamic state of the particle size distribution in the reactor, the interior of the reactor can be evenly divided into several areas by virtual grid division. Specifically, the software is used to construct a three-dimensional space model, and the internal space of the reactor is divided into multiple equal volume unit grids, each grid representing an area. At the same time, combined with the data input of the multi-point sensors arranged inside the reactor, the system automatically matches the monitoring range of each sensor with the grid unit position to ensure that each grid area has corresponding data coverage. In addition, the grid resolution can be dynamically adjusted by the software. For example, when the particle size changes in certain areas are large, the division density of the local grid is increased to enhance the fine monitoring capability. In this way, a virtual partition of the internal space of the reactor is formed, providing a basis for subsequent uniformity analysis.

[0062] Real-time detection of the uniformity of temperature, concentration and fluid flow rate distribution in each area inside the reactor requires multi-point sensor networks and data fusion technology. Specifically, high-precision sensors are arranged in each area inside the reactor to collect data on temperature, concentration and fluid flow rate. These sensors transmit real-time data to the central data processing system through the Industrial Internet of Things (IoT). The data processing system integrates the sensor data and quantifies the uniformity level of the data in each area through mathematical formulas (such as gradient analysis and standard deviation calculation), such as calculating the temperature gradient, concentration difference and the fluctuation range of the fluid flow rate. Combined with the dynamic threshold judgment mechanism, the system can monitor the uniformity of each area in real time. When any parameter in some areas exceeds the uniformity threshold, the particle size distribution uniformity analysis process is automatically triggered. At the same time, the system further improves the sensitivity and detection accuracy of uneven conditions through modeling of historical data and fitting analysis of real-time data. This method can dynamically and accurately capture the uneven distribution phenomenon in the reactor.

[0063] Explanation of "uniformity of temperature, concentration and fluid flow rate distribution": Temperature uniformity: refers to the degree of difference in temperature values ​​between regions. Large temperature differences may lead to significant changes in particle growth rates in different regions, affecting the consistency of particle size distribution. Concentration uniformity: refers to the consistency of distribution of solute concentration in various regions of the reactor. Uneven concentration will lead to unbalanced particle growth conditions and form particles that are too large or too small. Fluid flow rate uniformity: refers to whether the flow rate of the reaction liquid in each region of the reactor is uniform. Regions with too high or too low flow rates will cause local disturbances to particle growth and destroy the overall uniformity of particle size.

[0064] The core purpose of doing so is to solve the key technical problem of "local particle size distribution out of control" from the global perspective of the reactor. During the crystallization process of lithium hexafluorophosphate, due to the complexity and dynamics of the flow of the fluid inside the reactor, the distribution of temperature, concentration and flow rate in different areas is often uneven, and this unevenness directly leads to significant differences in the particle growth rate of local areas, thereby destroying the uniformity of particle size distribution. By dividing the area and detecting the uniformity of the temperature, concentration and fluid flow rate distribution in each area in real time, the abnormal state of local conditions can be accurately captured to ensure early identification of the risk of out-of-control particle size distribution. Compared with the limitations of the existing technology that relies on single-point data evaluation, this method based on regional division and uniformity detection can more comprehensively reflect the dynamic state of the entire reactor, thereby providing high-precision input for subsequent analysis and regulation. In addition, triggering the particle size distribution uniformity analysis process can realize automatic early warning and dynamic adjustment through software, effectively solving the problem that the existing technology cannot respond to the risk of out-of-control particle size distribution in a timely manner, thereby significantly improving the quality stability and production efficiency of lithium hexafluorophosphate products.

[0065] Real-time acquisition of crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate, and analysis after acquisition to generate the crystal growth driving force coefficient and particle uniformity distribution index of each area;

[0066] In this embodiment, the crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate is obtained in real time, and analyzed after acquisition to generate the crystal growth driving force coefficient and particle uniform distribution index of each area respectively, which specifically includes the following steps:

[0067] Real-time acquisition of crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate, and pre-processing after acquisition;

[0068] Real-time acquisition of crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate can be achieved by arranging a variety of high-precision sensor networks inside the reactor and combining them with the industrial Internet of Things (IoT). Specifically, first, temperature sensors, concentration sensors and flow rate sensors are evenly distributed in various areas inside the reactor to collect dynamic change data of temperature, concentration and flow rate in real time. The sensor packages the real-time collected data into data packets of small time periods through the data acquisition module and transmits them to the central control system through industrial communication protocols (such as Modbus and OPC-UA). After receiving these real-time data, the software module of the central control system marks them regionally, that is, each set of data corresponds to a specific area and type (such as temperature, concentration or flow rate). In addition, in order to ensure the stability and real-time nature of the data, the sensor sampling frequency and data transmission interval (for example, 10 times per second) can be set to meet the needs of dynamic monitoring during the crystallization of lithium hexafluorophosphate.

[0069] Extracting crystal growth dynamics information and particle distribution uniformity information from the pre-processed crystallization dynamics information of each region;

[0070] The core purpose of preprocessing is to clean and standardize the collected raw crystallization dynamic information to ensure the accuracy and reliability of subsequent analysis. First, since the sensor may be affected by environmental noise, electromagnetic interference or equipment accuracy, the collected data may contain outliers, missing values ​​or other invalid information, which will directly affect the results of parameter calculation. Therefore, the first step of preprocessing is data cleaning, including removing outliers by upper and lower limit screening methods and supplementing missing values ​​by interpolation. Secondly, the numerical range and unit of different types of data (such as temperature, concentration, flow rate) may vary greatly. In order to ensure that the contribution of each data to subsequent analysis is balanced, it is necessary to normalize it, such as using the maximum and minimum value normalization formula. Finally, in order to remove the interference of short-term fluctuations on the overall trend analysis, the sliding average filtering algorithm is added to smooth the data to enhance the continuity and reliability of dynamic information. The preprocessing process is automatically completed by the data processing module in the central control system. All processing rules are parameterized, and the preprocessing strategy can be dynamically adjusted to adapt to different production conditions.

[0071] The extracted crystal growth dynamics information and particle distribution uniformity information are analyzed to generate the crystal growth driving force coefficient and particle uniformity distribution index of each region respectively.

[0072] In this embodiment, the logic for obtaining the crystal growth driving force coefficient of each region is as follows:

[0073] The crystal growth dynamics information in the crystallization dynamics information of each pre-processed region is extracted, including the temperature gradient, solution saturation and fluid shear stress change rate of each region inside the reactor at different times during the crystallization process of lithium hexafluorophosphate, and calibrated as , and , represents the temperature gradient of the i-th region inside the reactor at time m during the crystallization of lithium hexafluorophosphate. It represents the solution saturation of the i-th region inside the reactor at time m during the crystallization process of lithium hexafluorophosphate. represents the rate of change of fluid shear stress in the i-th region of the reactor at time m during a period of time during the crystallization of lithium hexafluorophosphate, i=1, 2, 3, ..., k, m=1, 2, 3, ..., g, k and g are both positive integers;

[0074] During the crystallization process of lithium hexafluorophosphate, three types of data, namely, temperature gradient, solution saturation and fluid shear stress change rate, can be obtained in real time in each area of ​​the reactor through a multi-sensor network combined with an industrial Internet of Things (IoT) system, and processed and calibrated by software. Specifically, the temperature gradient can be obtained through high-precision temperature sensors distributed at different positions inside the reactor, and the temperature values ​​of each position are recorded in real time at each sampling moment. The temperature gradient is calculated by the software based on the temperature difference of adjacent measuring points; the solution saturation is obtained by the concentration sensor and the saturation concentration model. The concentration sensor collects the actual concentration of the solution in each area in real time. At the same time, the software calculates the corresponding saturation concentration based on the temperature in the area and the pre-established thermodynamic model, and then generates saturation data through the ratio of the two; the fluid shear stress change rate is obtained by the flow rate sensor and the pressure sensor. The flow rate sensor measures the velocity gradient of the fluid in real time, and the pressure sensor measures the change of the liquid shear force. The software fuses the flow rate and pressure data through the shear stress calculation formula based on fluid mechanics to obtain the real-time value of the shear stress, and further calculates the shear stress change rate through time difference. Among these three types of data, the temperature gradient reflects the change in thermal energy distribution, the solution saturation reflects the crystallization potential of the solute, and the fluid shear stress change rate represents the effect of fluid disturbance on the mass transfer of the crystal surface. Through the real-time collection, calibration and fusion of multi-sensor data by software, the dynamic state data of each area in the reactor can be generated efficiently and accurately, providing a reliable basis for subsequent calculation and analysis.

[0075] Calculate the crystal growth driving force coefficient of each region. The specific calculation formula is as follows:

[0076] ,

[0077] In the formula, CDC iis the crystal growth driving force coefficient of the ith region.

[0078] This calculation formula uses nonlinear calculations to accurately quantify the dynamic changes in the driving force for crystal growth by comprehensively considering three key factors: temperature gradient, solution saturation, and fluid shear stress change rate. The logarithmic function is used to amplify the combined effect of temperature gradient and solution saturation, and the shear stress change rate is introduced as a regulating factor to enhance the calculation sensitivity at low driving force. Exponential calculations are used to highlight the nonlinear amplification effect of shear stress change rate and temperature gradient on driving force, reflecting the potential exponential growth trend of driving force under high stress conditions; The square root operation is used to smooth the effect of saturation on the driving force, and a negative sign is introduced to simulate the possible limiting effect of solution supersaturation on crystal growth; the influence of outliers at a single moment is eliminated by averaging the time series, ensuring the stability and representativeness of the driving force evaluation. The overall formula can accurately reflect the multidimensional characteristics of the driving force for crystal growth in each region, providing a scientific basis for the subsequent particle size distribution evaluation.

[0079] The size of the crystal growth driving force coefficient of the i-th region directly affects the evaluation results of the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate, because this coefficient quantifies the potential dynamic differences of crystal growth in the region. When the values ​​of the crystal growth driving force coefficients of each region are close, it indicates that the temperature gradient, solution saturation, and fluid shear stress change rate inside the reactor are evenly distributed between different regions, and the crystal growth environment is highly consistent, thereby promoting the uniformity of the particle size distribution; conversely, when the driving force coefficients of different regions are significantly different, it indicates that there are large differences in the crystal growth conditions of each region, which can easily lead to asynchronous particle growth rates, thereby destroying the uniformity of the particle size distribution. Therefore, by calculating the crystal growth driving force coefficients of each region and analyzing its standard deviation between regions, the uniformity of the overall particle size distribution in the reactor can be quantified, providing a basis for subsequent control measures.

[0080] In this embodiment, the logic for obtaining the particle uniform distribution index of each region is as follows:

[0081] The particle distribution uniformity information in the crystallization dynamic information of each pre-processed area is extracted, including the concentration variance, flow velocity gradient and temperature fluctuation amplitude of each area inside the reactor at different times during the crystallization process of lithium hexafluorophosphate, and calibrated as , and , represents the concentration variance of the i-th region inside the reactor at time n during the crystallization process of lithium hexafluorophosphate. represents the flow velocity gradient of the i-th region inside the reactor at time n during the crystallization process of lithium hexafluorophosphate, represents the temperature fluctuation amplitude of the i-th region inside the reactor at time n during a period of time during the crystallization process of lithium hexafluorophosphate, i=1, 2, 3, ..., k, n=1, 2, 3, ..., h, k and h are both positive integers;

[0082] During the crystallization process of lithium hexafluorophosphate, three types of data, namely concentration variance, flow velocity gradient and temperature fluctuation amplitude, can be obtained in each area of ​​the reactor through the collaborative acquisition of multiple sensors and software data processing. The concentration variance can be collected in real time by the concentration sensors arranged in each area. The collected concentration data are statistically analyzed by the software, and the distribution consistency of the concentration is quantified by calculating the degree of deviation of these data from the average concentration value. The greater the degree of deviation, the higher the variance, indicating that the concentration distribution is more uneven; the flow velocity gradient measures the velocity value of the fluid at different positions in the reactor through the flow velocity sensor, and at the same time, combined with the spatial distribution of the sensor location, the software calculates the degree of velocity change between adjacent positions, thereby quantifying the consistency of fluid flow in the region. The greater the velocity change, the higher the flow velocity gradient, indicating that the fluid distribution is more uneven; the temperature fluctuation amplitude is calculated by the temperature data recorded by the high-precision temperature sensor at different times in each area. The software evaluates the temperature variation range by analyzing the difference between the highest and lowest temperature values ​​in the region. The larger the variation range, the higher the fluctuation amplitude, indicating that the thermal field distribution in the region is more unstable. Among these three types of data, concentration variance is used to reflect the discrete degree of solute distribution, velocity gradient is used to characterize the consistency of fluid flow, and temperature fluctuation amplitude is used to quantify the dynamic changes of thermal field distribution. Through the processing and analysis of these raw data by software, these three types of data can be efficiently obtained to provide accurate input for the subsequent calculation of particle uniform distribution index.

[0083] Calculate the particle uniform distribution index of each area. The specific calculation formula is as follows:

[0084] ,

[0085] Where, PDI i is the particle uniform distribution index of the ith region.

[0086] The calculation formula of the particle uniformity index in each area comprehensively quantifies the particle distribution uniformity through nonlinear operations on three key data types: concentration variance, flow velocity gradient, and temperature fluctuation amplitude, accurately reflecting the impact of the conditions in each area on the uniformity. The relationship between concentration variance and velocity gradient is described by using negative exponential operation. When the concentration variance is large or the velocity gradient is small, the uniformity decreases sharply, while the high velocity gradient can significantly alleviate the negative impact of concentration difference. The influence of temperature fluctuation on uniformity is magnified by cubic operation, emphasizing that the more severe the temperature fluctuation, the greater the damage to uniformity; Logarithmic calculations are used to quantify the ratio of velocity gradient to temperature fluctuation amplitude, reflecting how the stability of the velocity gradient improves uniformity under low temperature fluctuation conditions. The formula finally processes the dynamic data at different times through time averaging to ensure that the uniform distribution index can accurately reflect the overall distribution trend over a long period of time. This calculation method takes into account the direct and indirect effects of various factors on the uniformity of particle distribution, providing a scientific basis for subsequent uniformity evaluation.

[0087] The size of the particle uniform distribution index of the i-th region directly reflects the uniformity of the particle distribution conditions in the region, and is closely related to the evaluation of the overall uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate. When the particle uniform distribution index values ​​of each region are low and close to each other, it means that the concentration variance is small, the flow velocity gradient is stable, and the temperature fluctuation amplitude is low. The particle growth environment in each region is relatively consistent, and the overall particle size distribution of the reactor tends to be uniform; on the contrary, when the particle uniform distribution index of different regions is significantly different or the index value of some regions is high, it indicates that the local concentration distribution is uneven, the flow velocity fluctuates violently, or the temperature is unstable. This inconsistency will lead to a decrease in the overall uniformity of the particle size distribution. By calculating and analyzing the particle uniform distribution index of each region and its standard deviation among all regions, the uniformity of the particle size distribution inside the reactor can be comprehensively evaluated, providing a scientific basis for dynamic regulation.

[0088] A particle size distribution evaluation model is constructed based on the crystal growth driving force coefficient and particle uniformity distribution index of each generated region, and the particle size distribution coefficient of each region is generated. After the generation, a comprehensive analysis is performed to generate an evaluation coefficient;

[0089] In this embodiment, the crystal growth driving force coefficient CDC of each region is generated. i and particle distribution index PDI i Construct a particle size distribution evaluation model and generate the particle size distribution coefficient of each area by weighted summation according to the formula:

[0090] ,

[0091] In the formula, GDC i is the particle size distribution coefficient of the i-th region, ω1 and ω2 are the crystal growth driving force coefficients CDC of each region respectively i and particle distribution index PDI i The non-zero weight coefficient of , and ω1+ω2=1;

[0092] This part can be realized by performing a weighted summation operation on the generated crystal growth driving force coefficient and particle uniform distribution index through the software module. Specifically, first, the software pre-sets the weight coefficients ω1 and ω2 according to the experimental data and process requirements. The setting of the weight coefficients needs to be based on the crystal growth driving force coefficient CDC. i and particle distribution index PDI i The weight of the particle size distribution is dynamically adjusted according to the degree of influence on the uniformity of the particle size distribution. For example, when the change in the driving force of crystal growth has a more significant impact on the particle size distribution, a higher ω1 value (such as 0.7) can be set, while ω2 is lower (such as 0.3); conversely, when the change in the particle uniformity index has a dominant effect on the particle size distribution, the weight of ω2 is increased. The weight satisfies the constraint condition of ω1+ω2=1 to ensure that the total effect of the weighted sum is consistent. When calculating, the software calculates the weight of the particle size distribution by ω1*CDC for each area. i and ω2*PDI i Sum each item to generate the corresponding particle size distribution coefficient GDC i This weighted summation approach allows for differentiating the importance of different factors when evaluating the uniformity of particle size distribution, thereby achieving more accurate regional assessment results.

[0093] The particle size distribution coefficient GDC of each area i The evaluation coefficient ENC is generated by calculating its standard deviation according to the formula: .

[0094] The generated evaluation coefficients are analyzed to evaluate the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate, and the uniformity of the particle size distribution is divided into high uniformity, normal uniformity and non-uniformity according to the evaluation results, and a high uniformity signal, a normal uniformity signal and a non-uniform signal are generated respectively;

[0095] In this embodiment, the predetermined evaluation coefficient threshold interval [ENC min , ENC max ], and after determination, it is compared with the generated evaluation coefficient ENC, and the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate is evaluated according to the comparison results, and the uniformity of the particle size distribution is divided into high uniformity, normal uniformity and non-uniformity according to the evaluation results, and a high uniformity signal, a normal uniformity signal and a non-uniform signal are generated respectively. The specific comparison analysis is as follows:

[0096] If ENC<ENC min , during the crystallization of lithium hexafluorophosphate, the uniformity of the particle size distribution inside the reactor is high, generating a high uniformity signal;

[0097] This indicates that the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate is extremely uniform. At this time, the crystal growth conditions (such as temperature, concentration and flow rate) in each area of ​​the reactor are relatively consistent, the particle growth rate and distribution state tend to be balanced, and the particle size is stable and uniformly distributed. Such high uniformity means that the quality of lithium hexafluorophosphate products can reach the optimal level, which is conducive to meeting the strict requirements of high-performance lithium batteries for conductivity and purity, and can significantly reduce the difficulty of subsequent separation and purification processes, thereby improving production efficiency and reducing costs.

[0098] If ENC min ≤ENC≤ENC max , during the crystallization of lithium hexafluorophosphate, the uniformity of the particle size distribution inside the reactor is normal, and a normal uniform signal is generated;

[0099] This indicates that the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate is within the normal uniform range. At this time, although there are certain differences in the crystal growth conditions in each area, these differences are still within an acceptable range and have limited impact on the overall particle size distribution of the particles. Such normal uniformity indicates that the product quality can basically meet the requirements of lithium battery applications, but may be slightly lower than the optimal state. At the same time, in the subsequent separation and purification steps, it is still necessary to avoid possible slight particle aggregation to ensure product consistency.

[0100] If ENC>ENC max ,During the crystallization process of lithium hexafluorophosphate, the uniformity of the particle size distribution inside the reactor is uneven, generating an uneven signal.

[0101] This situation shows that there is significant non-uniformity in the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate. At this time, the crystal growth conditions in different areas of the reactor vary greatly, which may lead to the coexistence of particles that are too large or too small, and the quality and stability of the product are seriously affected. Such non-uniformity will directly reduce the purity and particle size consistency of lithium hexafluorophosphate, which may lead to a decrease in the conductivity of lithium batteries and even increase the difficulty and cost of subsequent separation and purification processes. At this time, it is necessary to take regulatory measures quickly to adjust the process parameters to restore the uniformity of particle size distribution.

[0102] Determine that the pre-set evaluation coefficient threshold interval can be dynamically adjusted automatically by the software through methods based on historical data analysis and experimental verification. Specifically, first, the software statistically analyzes the evaluation coefficient data of multiple batches of lithium hexafluorophosphate crystallization processes in the past to identify the distribution characteristics of the evaluation coefficients corresponding to high uniformity, normal uniformity, and unevenness. The software calculates the mean and standard deviation of these historical data, divides the overall distribution range of the evaluation coefficient into high uniformity areas, normal uniformity areas, and uneven areas, and preliminarily sets the minimum and maximum values ​​of the threshold interval. Then, the interval is further optimized through experimental verification. For example, in the actual process, the particle size distribution quality results corresponding to different evaluation coefficient ranges are compared with the battery performance requirements, and the software iteratively adjusts the boundary value of the threshold interval to ensure that the threshold interval accurately reflects the uniformity of the particle size distribution in actual production. Finally, the software stores the determined threshold interval in a parameterized manner and supports dynamic updates to adapt to changes in different production batches or process conditions. In this way, it can be ensured that the determination of the evaluation coefficient threshold interval has both a scientific basis and can flexibly adapt to actual production needs.

[0103] In the case of generating uneven signals, continuously obtain several evaluation coefficients generated subsequently, conduct comprehensive analysis, generate different levels of early warning signals of particle size distribution out of control risk, and take corresponding control measures for different early warning levels;

[0104] In this embodiment, when an uneven signal is generated, several evaluation coefficients generated subsequently are continuously obtained and calibrated as ENC x , x represents the numbers of several evaluation coefficients subsequently generated when an uneven signal is generated, x=1, 2, 3, ..., d, d is a positive integer;

[0105] In the case of generating uneven signals, the continuous acquisition of several evaluation coefficients generated subsequently can be achieved through the real-time monitoring of the sensor network and the collaborative data processing module of the software. Specifically, the sensor network is distributed in various areas inside the reactor, and key data such as temperature, concentration and flow rate are collected in real time, which are transmitted to the monitoring module of the central control system. The software first pre-processes the sensor data, including outlier removal, noise filtering and data normalization, to ensure the quality and reliability of the original data. Subsequently, the software calculates the crystal growth driving force coefficient and particle uniformity distribution index of each area in real time according to the preset calculation rules, and generates the current evaluation coefficient by constructing a particle size distribution evaluation model. For several evaluation coefficients generated subsequently, the software continuously records the evaluation coefficients calculated at each moment and assigns them numbers through the dynamic storage and update mechanism of time series data. At the same time, the software monitors the changing trends of these evaluation coefficients in real time, and compares the newly generated evaluation coefficients with the stored historical data to identify changes in process status and generate real-time feedback. In this way, the software can efficiently and continuously obtain the evaluation coefficients, providing reliable basic data for subsequent comprehensive analysis and regulation.

[0106] Calculate several evaluation coefficients ENC generated subsequently x The average ENC — , according to the formula: ;

[0107] Calculate several evaluation coefficients ENC generated subsequently x Standard deviation of ENC σ , according to the formula: ;

[0108] Determine a preset mean value threshold of a number of evaluation coefficients subsequently generated in the case of generating an inhomogeneous signal and preset standard deviation threshold , the average value of several evaluation coefficients ENC — and standard deviation ENC σ The preset average value thresholds of several evaluation coefficients are and preset standard deviation threshold Compare and generate different levels of early warning signals for the risk of particle size distribution out of control according to the comparison results, and take corresponding control measures for different early warning levels. The specific comparison and analysis are as follows:

[0109] like and , generating a high-level early warning signal for the risk of out-of-control particle size distribution. The control measures taken are: optimizing the fluid flow state by adjusting the stirring speed and the position of the stirring paddle to reduce the flow rate fluctuation, reducing the temperature gradient inside the reactor to reduce the unevenness of the thermal field distribution, and improving the uniformity of the solute distribution by adjusting the solution feed rate and concentration distribution. Real-time monitoring of the temperature, concentration and flow rate changes after control, and dynamically adjusting the control parameters according to the feedback results until the particle size distribution returns to a uniform state;

[0110] This situation shows that the particle size distribution inside the reactor is not only significantly uneven (high standard deviation), but also the overall trend continues to deteriorate (average value is high). At this time, the core process parameters such as temperature, flow rate and solute concentration in the reactor vary significantly between different areas, resulting in excessively fast or slow particle growth in local areas and out-of-control of the overall particle size distribution. The impacts include: a significant decline in the quality of lithium hexafluorophosphate products, insufficient particle size consistency may lead to unstable battery performance, and increase the complexity and cost of subsequent separation and purification processes.

[0111] The software can collect temperature, flow rate and concentration data of each area of ​​the reactor in real time, and set dynamic adjustment strategies in combination with historical data. Specifically, first, the software analyzes the temperature gradient of each area in the reactor. If the temperature in some areas is too high, the cooling device is controlled to cool these areas; for areas with low temperatures, the heat source is adjusted to heat them to reduce the unevenness of the temperature gradient. Secondly, the software optimizes the flow rate data, balances the fluid flow state and reduces flow rate fluctuations by changing the speed and angle of the agitator. Finally, combined with the solute concentration data, the concentration and feed rate of the solution in the feeding system are adjusted to ensure uniform distribution of the solute. Through these measures, the uniformity of the particle size distribution inside the reactor is gradually restored to ensure stable product quality.

[0112] like and , generating an early warning signal for the risk of loss of control of the medium-level particle size distribution. The control measures taken are: gradually adjusting the temperature distribution of each area inside the reactor to reduce temperature fluctuations, balancing the flow rate difference by optimizing the stirring speed and changing the fluid direction, alleviating the non-uniformity of the solute concentration by compensation, and tracking the parameter changes of each area after control in real time. According to the monitoring results, the control strategy is optimized to ensure that the uniformity of the particle size distribution inside the reactor is gradually improved;

[0113] This indicates that although the overall particle size distribution trend has improved (lower average), there is still significant heterogeneity in local areas (higher standard deviation). The warning signal for the risk of loss of control of the medium-level particle size distribution indicates that the overall trend of the particle size distribution inside the reactor is good, but there is still significant heterogeneity in the particle size distribution in local areas. It may be due to large fluctuations in temperature, flow rate or concentration in certain local areas, resulting in significant deviations in particle size in these areas. The impacts include: the overall quality of the product is threatened by local defects, and small areas of non-compliance may appear, which will reduce product consistency and increase the complexity of production adjustments.

[0114] The software identifies local areas with large fluctuations through regional analysis. For areas with obvious temperature fluctuations, the software controls local heating or cooling equipment to adjust the temperature in a small range, gradually reducing the temperature difference between regions; for areas with large flow rate changes, the software achieves flow rate balance by adjusting the speed of the local agitator or changing the direction of the fluid; for areas with uneven concentration, the software controls the local solution compensation system to increase or decrease the solute supply and alleviate the concentration difference. The software monitors the control effect in real time and dynamically optimizes the control parameters to ensure that local problems do not spread and gradually improve the consistency of particle size distribution.

[0115] like and , generating a warning signal for the risk of low-level particle size distribution being out of control, and taking the following control measures: reducing the saturation of the overall solution in the reactor to reduce the crystal growth rate, reducing the fluid shear stress by slowing down the overall stirring speed to avoid growth disturbances, and reducing the impact of temperature changes on particle size distribution by adjusting the temperature distribution control range. At the same time, continuously monitoring the state inside the reactor, recording changes in particle size distribution, and ensuring that the overall uniformity is gradually improved;

[0116] This indicates that the overall particle size distribution trend has deteriorated (the average value is high), but the difference between regions is not large (the standard deviation is low). The low-level particle size distribution out-of-control risk warning signal indicates that the overall trend of the particle size distribution in the reactor has deteriorated, but the difference between local regions is small, that is, the distribution deviation is not large but the overall uniformity level has decreased. The impact is mainly manifested in that the product quality begins to deviate from the optimal state, but is still within an acceptable range. If it is not adjusted in time, it may lead to a gradual deterioration of uniformity, which in turn affects the stability of the product and subsequent process operations.

[0117] The software analyzes the overall trend changes and makes gentle adjustments to the global parameters. First, the software controls the saturation of the solution, reduces the crystal growth rate by reducing the solute concentration in the solution, and makes the particle growth tend to be synchronous; second, the software slightly slows down the stirring speed to reduce the interference of fluid shear stress on particle growth; finally, by adjusting the power of the heat source in a small range, the impact of temperature fluctuations on overall uniformity is reduced. During the control process, the software continuously monitors the changing trend of the particle size distribution to ensure that the adjustment of measures can gradually improve the distribution uniformity and avoid further deterioration of the problem.

[0118] like and , generate a warning signal of no risk of loss of control and do not take any regulatory measures.

[0119] This situation shows that the particle size distribution in the reactor has gradually tended to a uniform state. The absence of a risk of loss of control warning signal indicates that the uniformity of the particle size distribution inside the reactor has reached the optimal state, and the overall trend and local consistency are at a good level, which meets the requirements of high-quality lithium hexafluorophosphate products. In this state, no additional adjustments are required, which can effectively reduce process energy consumption while ensuring production stability and efficiency. In this case, the software does not need to be regulated, but only needs to continue to maintain real-time monitoring to ensure the stability of the uniformity state. Specifically, the software continuously collects temperature, flow rate and concentration data, records all parameters of the current production status, and stores them as historical benchmark values ​​for reference in subsequent batch production; at the same time, the software regularly analyzes monitoring data to verify the continued stability of the process and ensure that no new abnormal fluctuations occur. In this way, reliable data can be accumulated for future process optimization while maintaining efficient production.

[0120] The preset mean value threshold and the preset standard deviation threshold of several evaluation coefficients generated subsequently in the case of generating non-uniform signals can be determined by the software in combination with historical data analysis and dynamic parameter optimization. Specifically, the software first extracts the evaluation coefficient data after generating non-uniform signals from historical production batches, calculates the mean value and standard deviation of these evaluation coefficients through statistical analysis, and calibrates the process state corresponding to these parameters based on product quality requirements. For example, by analyzing the evaluation coefficient range corresponding to high-quality and low-quality products in historical data, the boundaries of the mean value and standard deviation of the evaluation coefficient required for high uniformity are determined, and the initial threshold is generated. Then, the software dynamically optimizes the actual production data of the current batch, and adjusts the preset threshold range of the mean value and standard deviation by comparing the deviation between the real-time evaluation coefficient and the historical benchmark value to adapt it to the current process conditions. The software also verifies through simulation experiments whether the set threshold can effectively distinguish different levels of uniformity, and iteratively updates the threshold range when necessary. Finally, the software stores the determined threshold interval as a dynamic parameter for continuous monitoring and risk assessment to achieve precise process control and tuning.

[0121] The reason why several subsequent evaluation coefficients need to be continuously obtained when an uneven signal is generated, and their average values ​​and standard deviations are calculated for comparison, is to dynamically and accurately evaluate the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate and its changing trend. This method can minimize the impact of the accidental or instantaneous fluctuations of a single generated uneven signal, and provide a more stable and comprehensive uniformity evaluation result by comprehensively analyzing the data trends of multiple groups of evaluation coefficients. At the same time, combined with the preset average value and standard deviation thresholds, it is not only possible to quantitatively judge the state of the current particle size distribution, but also to divide the risk level according to the degree of deviation and generate corresponding early warning signals. By taking differentiated control measures for high, medium and low level early warning signals, it is possible to effectively intervene in different degrees of unevenness problems, which not only avoids the waste of resources caused by excessive control, but also prevents product quality problems that may be caused by insufficient control, ensuring the efficiency of the production process and the stability and consistency of lithium hexafluorophosphate products.

[0122] The crystallization dynamic information and particle size distribution coefficient of each area inside the reactor are continuously monitored, and the parameters and control strategies of the particle size distribution evaluation model are dynamically adjusted according to the real-time monitoring results. At the same time, historical data are stored and analyzed to optimize the model parameters and control strategies.

[0123] The crystallization dynamic information and particle size distribution coefficient of each area inside the reactor are continuously monitored, and the parameters and control strategies of the particle size distribution evaluation model are dynamically adjusted. At the same time, historical data are stored and analyzed. This process is to ensure that the uniformity of the particle size distribution in the production process can be optimized in real time as the process conditions change, and to prevent quality problems caused by external disturbances or equipment fluctuations. Through real-time monitoring, it is possible to capture subtle changes in dynamic data such as temperature, concentration and flow rate, and generate the latest particle size distribution coefficient. The software can identify potential process deviations in a timely manner based on these data. In terms of implementation, the software collects dynamic information from each area of ​​the reactor in real time by connecting to a high-precision sensor network, and uses the data processing module to pre-process the information, including noise removal, normalization and outlier filtering. Subsequently, the software dynamically adjusts the core parameters of the particle size distribution evaluation model (such as weight coefficients, threshold intervals, etc.) in combination with the latest data to optimize the model's adaptability to the current process status. The control strategy is automatically updated according to the adjusted evaluation results to ensure that the control measures are accurate and efficient. At the same time, the software stores all monitoring data and adjustment records, and regularly analyzes historical data to identify long-term trends and repetitive problems, providing a scientific basis for further optimization of model parameters and control strategies. This closed-loop feedback mechanism ensures the flexibility and stability of process control, thereby improving production efficiency and ensuring the consistency and high quality of lithium hexafluorophosphate products.

[0124] In order to verify the effectiveness of the dynamic adjustment measures and particle size distribution evaluation model proposed in the present invention, multiple groups of experiments were conducted to evaluate the effects of different technical conditions on particle size distribution uniformity and product quality. In the experiment, for the crystallization process of lithium hexafluorophosphate, "undivided area", "divided into five areas" and "combined with dynamic adjustment and historical optimization" were selected as some experimental examples. By monitoring parameters such as the crystal growth driving force coefficient, particle uniformity distribution index and particle size distribution coefficient in each area, the effects of various technical conditions on particle size uniformity were comprehensively analyzed. The following experimental data table compares the technical parameters and effects under different experimental conditions, fully illustrating the significant advantages of the technical solution of the present invention in improving particle size distribution uniformity and product quality.

[0125] Table 1 Experimental data for evaluating the uniformity of particle size distribution during the crystallization of lithium hexafluorophosphate

[0126]

[0127] The experimental data table clearly shows the impact of different technical conditions on the particle size distribution uniformity and quality indicators during the crystallization of lithium hexafluorophosphate. The following is a detailed interpretation of each experimental condition:

[0128] No dynamic adjustment, no partition (A1): Without regional division and dynamic adjustment measures, the experiment relies only on single-point data for monitoring, which cannot reflect the dynamic changes of various areas in the reactor. Due to the inability to capture the heterogeneity of local conditions, the values ​​of the crystal growth driving force coefficient (CDC) and the particle uniformity index (PDI) are high, the particle size distribution evaluation coefficient (GDC) fluctuates significantly, and the final evaluation coefficient (ENC) reaches 1.2. The particle size uniformity level is "uneven", and the quality index compliance rate is only 78.5%, which is far below the process requirements. This result shows that without dynamic adjustment measures and partition monitoring, the particle size distribution is difficult to effectively control, and the product quality is seriously affected.

[0129] Implement dynamic adjustment and zone monitoring (A2): Under the experimental conditions divided into five zones, the driving force of crystal growth and the consistency of uniform particle distribution were significantly improved by monitoring and analyzing the dynamic data of each zone. The tabular data show that the CDC and PDI values ​​of each zone are relatively close, the particle size distribution coefficient (GDC) is relatively stable, the evaluation coefficient (ENC) is reduced to 0.8, the particle size uniformity level is improved to "normal uniformity", and the quality index compliance rate is increased to 92.0%. This result shows that by dividing multiple zones and combining dynamic adjustment measures, the uniformity of particle size distribution can be significantly improved and product quality can be improved.

[0130] Implement dynamic adjustment and combine historical optimization (A3): Under the experimental conditions of further combining dynamic adjustment measures and historical data optimization, the parameters of each area are further optimized. The data show that the CDC and PDI values ​​of each area are closer, the particle size distribution coefficient (GDC) shows a high degree of consistency, the evaluation coefficient (ENC) drops to 0.3, the particle size uniformity level reaches "high uniformity", and the quality index compliance rate is increased to 99.5%. This result fully proves that the real-time feedback and historical optimization of dynamic adjustment measures can more effectively improve the uniformity of particle size distribution, providing a strong guarantee for the stability of product performance.

[0131] Through comparative analysis of experimental data, dynamic adjustment measures have shown their core role in optimizing particle size distribution uniformity, which is specifically reflected in the following aspects:

[0132] Refined expansion of monitoring scope: Dynamic adjustment measures refine the ability to capture local dynamic changes inside the reactor through real-time monitoring technology of partitions, so that areas that cannot be covered by traditional single-point monitoring methods can be fully monitored. This multi-point monitoring provides an accurate data basis for subsequent adjustments, especially in the fluctuation areas of key parameters such as temperature, concentration and flow rate. Dynamic monitoring reveals local out-of-control problems that have been ignored before, significantly improving monitoring accuracy.

[0133] Real-time feedback and precise optimization of control strategies: Through dynamic adjustment measures, the system can respond to changes in the internal parameters of the reactor in real time, and combine the particle size distribution evaluation model to accurately adjust key process parameters such as stirring speed, temperature control range and feed concentration. This fast feedback mechanism not only effectively alleviates the risk of local loss of control, but also significantly reduces the particle size distribution coefficient (GDC) and evaluation coefficient (ENC), thereby achieving a high degree of consistency in particle size distribution. In addition, the optimization strategy combined with historical data further improves the adaptability and control effect of the model, making process regulation more stable and efficient.

[0134] Comprehensive guarantee of product quality: Experimental results show that dynamic adjustment measures can reduce the evaluation coefficient (ENC) to the optimal state (0.3) under high uniformity conditions, significantly improving the product quality index compliance rate to 99.5%. This highly uniform particle size distribution is crucial to improving the purity, stability and conductivity of lithium hexafluorophosphate products, while simplifying the subsequent separation and purification processes and reducing overall production costs. This fully proves that dynamic adjustment measures not only solve the core problem of out-of-control particle size distribution, but also provide reliable guarantees for the high performance and high stability of lithium battery materials.

[0135] In general, the dynamic adjustment measures have improved the uniformity of particle size distribution to the industry-leading level through detailed monitoring, precise control and efficient feedback optimization, laying a solid foundation for the industrial production of lithium hexafluorophosphate. This technical solution is highly adaptable and scalable, fully demonstrating its application potential in complex production scenarios.

[0136] like Figure 2 The lithium hexafluorophosphate production control system shown includes a region division and uniformity detection module, a crystallization dynamic information acquisition and parameter generation module, a particle size distribution evaluation and coefficient generation module, a uniformity evaluation and signal generation module, a loss of control risk warning and regulation module, and a real-time monitoring and model optimization module;

[0137] The regional division and uniformity detection module divides the interior of the reactor used for lithium hexafluorophosphate crystallization into several regions during the crystallization process of lithium hexafluorophosphate, and performs real-time detection of the temperature, concentration and fluid flow rate distribution uniformity of each region inside the reactor. When non-uniformity is detected in certain regions, the particle size distribution uniformity analysis process is triggered;

[0138] The crystallization dynamic information acquisition and parameter generation module acquires the crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate in real time, analyzes it after acquisition, and generates the crystal growth driving force coefficient and particle uniform distribution index of each area respectively;

[0139] The particle size distribution evaluation and coefficient generation module builds a particle size distribution evaluation model based on the crystal growth driving force coefficient and particle uniform distribution index of each region, generates the particle size distribution coefficient of each region, and performs a comprehensive analysis after the generation to generate the evaluation coefficient;

[0140] The uniformity evaluation and signal generation module analyzes the generated evaluation coefficients, evaluates the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate, and divides the uniformity of the particle size distribution into high uniformity, normal uniformity and non-uniformity according to the evaluation results, and generates high uniformity signals, normal uniformity signals and non-uniform signals respectively;

[0141] The out-of-control risk warning and control module, when generating uneven signals, continuously obtains several evaluation coefficients generated subsequently, conducts comprehensive analysis, generates different levels of out-of-control risk warning signals for particle size distribution, and takes corresponding control measures for different warning levels;

[0142] The real-time monitoring and model optimization module continuously monitors the crystallization dynamic information and particle size distribution coefficient of each area inside the reactor, and dynamically adjusts the parameters and control strategies of the particle size distribution evaluation model according to the real-time monitoring results. At the same time, it stores and analyzes historical data to optimize model parameters and control strategies.

[0143] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0144] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center by wired or wireless (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a tape), an optical medium (for example, a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0145] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0146] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0147] In the several embodiments provided in the present application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0148] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0149] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0150] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A lithium hexafluorophosphate production control method, characterized in that: The specific steps include: During the crystallization process of lithium hexafluorophosphate, the interior of the reactor used for lithium hexafluorophosphate crystallization is evenly divided into several areas, and the temperature, concentration and fluid flow rate distribution uniformity of each area inside the reactor are detected in real time. When uneven conditions are detected in some areas, the particle size distribution uniformity analysis process is triggered; Real-time acquisition of crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate, and analysis after acquisition to generate the crystal growth driving force coefficient and particle uniformity distribution index of each area; A particle size distribution evaluation model is constructed based on the crystal growth driving force coefficient and particle uniformity distribution index of each generated region, and the particle size distribution coefficient of each region is generated. After the generation, a comprehensive analysis is performed to generate an evaluation coefficient; The generated evaluation coefficients are analyzed to evaluate the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate, and the uniformity of the particle size distribution is divided into high uniformity, normal uniformity and non-uniformity according to the evaluation results, and a high uniformity signal, a normal uniformity signal and a non-uniform signal are generated respectively; In the case of generating uneven signals, continuously obtain several evaluation coefficients generated subsequently, conduct comprehensive analysis, generate different levels of early warning signals of particle size distribution out of control risk, and take corresponding control measures for different early warning levels; The crystallization dynamic information and particle size distribution coefficient of each area inside the reactor are continuously monitored, and the parameters and control strategies of the particle size distribution evaluation model are dynamically adjusted according to the real-time monitoring results. At the same time, historical data are stored and analyzed to optimize the model parameters and control strategies.

2. The lithium hexafluorophosphate production control method according to claim 1, characterized in that: The crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate is obtained in real time, and analyzed after acquisition to generate the crystal growth driving force coefficient and particle uniform distribution index of each area respectively, which specifically includes the following steps: Real-time acquisition of crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate, and pre-processing after acquisition; Extracting crystal growth dynamics information and particle distribution uniformity information from the pre-processed crystallization dynamics information of each region; The extracted crystal growth dynamics information and particle distribution uniformity information are analyzed to generate the crystal growth driving force coefficient and particle uniformity distribution index of each region respectively.

3. The lithium hexafluorophosphate production control method according to claim 2, characterized in that: The logic for obtaining the crystal growth driving force coefficient of each region is as follows: The crystal growth dynamics information in the crystallization dynamics information of each pre-processed region is extracted, including the temperature gradient, solution saturation and fluid shear stress change rate of each region inside the reactor at different times during the crystallization process of lithium hexafluorophosphate, and calibrated as , and , represents the temperature gradient of the i-th region inside the reactor at time m during the crystallization of lithium hexafluorophosphate. It represents the solution saturation of the i-th region inside the reactor at time m during the crystallization process of lithium hexafluorophosphate. represents the rate of change of fluid shear stress in the i-th region of the reactor at time m during a period of time during the crystallization of lithium hexafluorophosphate, i=1, 2, 3, ..., k, m=1, 2, 3, ..., g, k and g are both positive integers; Calculate the crystal growth driving force coefficient of each region. The specific calculation formula is as follows: ， In the formula, CDC i is the crystal growth driving force coefficient of the ith region.

4. The lithium hexafluorophosphate production control method according to claim 3, characterized in that: The logic for obtaining the particle uniform distribution index of each region is as follows: The particle distribution uniformity information in the crystallization dynamic information of each pre-processed area is extracted, including the concentration variance, flow velocity gradient and temperature fluctuation amplitude of each area inside the reactor at different times during the crystallization process of lithium hexafluorophosphate, and calibrated as , and , represents the concentration variance of the i-th region inside the reactor at time n during the crystallization process of lithium hexafluorophosphate. represents the flow velocity gradient of the i-th region inside the reactor at time n during the crystallization process of lithium hexafluorophosphate, represents the temperature fluctuation amplitude of the i-th region inside the reactor at time n during a period of time during the crystallization process of lithium hexafluorophosphate, i=1, 2, 3, ..., k, n=1, 2, 3, ..., h, k and h are both positive integers; Calculate the particle uniform distribution index of each area. The specific calculation formula is as follows: ， Where, PDI i is the particle uniform distribution index of the ith region.

5. The lithium hexafluorophosphate production control method according to claim 4, characterized in that: The crystal growth driving force coefficient CDC for each generated region i and particle distribution index PDI i Construct a particle size distribution evaluation model and generate the particle size distribution coefficient of each area by weighted summation according to the formula: ， In the formula, GDC i is the particle size distribution coefficient of the ith region, ω1 and ω2 are the crystal growth driving force coefficients CDC of each region respectively i and particle distribution index PDI i The non-zero weight coefficient of , and ω1+ω2=1; The particle size distribution coefficient GDC of each area i The evaluation coefficient ENC is generated by calculating its standard deviation according to the formula: .

6. The lithium hexafluorophosphate production control method according to claim 5, characterized in that: Determine the pre-set evaluation coefficient threshold interval [ENC min , ENC max ], and after determination, it is compared with the generated evaluation coefficient ENC, and the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate is evaluated according to the comparison results, and the uniformity of the particle size distribution is divided into high uniformity, normal uniformity and non-uniformity according to the evaluation results, and a high uniformity signal, a normal uniformity signal and a non-uniform signal are generated respectively. The specific comparison analysis is as follows: If ENC<ENC min , during the crystallization of lithium hexafluorophosphate, the uniformity of the particle size distribution inside the reactor is high, generating a high uniformity signal; If ENC min ≤ENC≤ENC max , during the crystallization of lithium hexafluorophosphate, the uniformity of the particle size distribution inside the reactor is normal, and a normal uniform signal is generated; If ENC>ENC max ,During the crystallization process of lithium hexafluorophosphate, the uniformity of the particle size distribution inside the reactor is uneven, generating an uneven signal.

7. The lithium hexafluorophosphate production control method according to claim 6, characterized in that: In the case of generating an uneven signal, several evaluation coefficients generated subsequently are continuously obtained and calibrated as ENC x , x represents the numbers of several evaluation coefficients subsequently generated when an uneven signal is generated, x=1, 2, 3, ..., d, d is a positive integer; Calculate the subsequent generated evaluation coefficients ENC x The average ENC — , according to the formula: ; Calculate the subsequent generated evaluation coefficients ENC x Standard deviation of ENC σ , according to the formula: ； Determine a preset mean value threshold of a number of evaluation coefficients subsequently generated in the case of generating an inhomogeneous signal and preset standard deviation threshold , the average value of several evaluation coefficients ENC — and standard deviation ENC σ The preset average value thresholds of several evaluation coefficients are and preset standard deviation threshold Compare and generate different levels of early warning signals for the risk of particle size distribution out of control according to the comparison results, and take corresponding control measures for different early warning levels. The specific comparison and analysis are as follows: like and , generating a high-level early warning signal for the risk of out-of-control particle size distribution. The control measures taken are: optimizing the fluid flow state by adjusting the stirring speed and the position of the stirring paddle to reduce the flow rate fluctuation, reducing the temperature gradient inside the reactor to reduce the unevenness of the thermal field distribution, and improving the uniformity of the solute distribution by adjusting the solution feed rate and concentration distribution. Real-time monitoring of the temperature, concentration and flow rate changes after control, and dynamically adjusting the control parameters according to the feedback results until the particle size distribution returns to a uniform state; like and , generating an early warning signal for the risk of loss of control of the medium-level particle size distribution. The control measures taken are: gradually adjusting the temperature distribution of each area inside the reactor to reduce temperature fluctuations, balancing the flow rate difference by optimizing the stirring speed and changing the fluid direction, alleviating the non-uniformity of the solute concentration by compensation, and tracking the parameter changes of each area after control in real time. According to the monitoring results, the control strategy is optimized to ensure that the uniformity of the particle size distribution inside the reactor is gradually improved; like and , generating a warning signal for the risk of low-level particle size distribution being out of control, and taking the following control measures: reducing the saturation of the overall solution in the reactor to reduce the crystal growth rate, reducing the fluid shear stress by slowing down the overall stirring speed to avoid growth disturbances, and reducing the impact of temperature changes on particle size distribution by adjusting the temperature distribution control range. At the same time, continuously monitoring the state inside the reactor, recording changes in particle size distribution, and ensuring that the overall uniformity is gradually improved; like and , generate a warning signal of no risk of loss of control and do not take any regulatory measures.

8. A lithium hexafluorophosphate production control system, which is used to implement the lithium hexafluorophosphate production control method according to any one of claims 1 to 7, characterized in that: It includes the module of area division and uniformity detection, the module of crystallization dynamic information acquisition and parameter generation, the module of particle size distribution evaluation and coefficient generation, the module of uniformity evaluation and signal generation, the module of out-of-control risk warning and regulation, and the module of real-time monitoring and model optimization; The regional division and uniformity detection module divides the interior of the reactor used for lithium hexafluorophosphate crystallization into several regions during the crystallization process of lithium hexafluorophosphate, and performs real-time detection of the temperature, concentration and fluid flow rate distribution uniformity of each region inside the reactor. When non-uniformity is detected in certain regions, the particle size distribution uniformity analysis process is triggered; The crystallization dynamic information acquisition and parameter generation module acquires the crystallization dynamic information of each area inside the reactor during the crystallization of lithium hexafluorophosphate in real time, analyzes it after acquisition, and generates the crystal growth driving force coefficient and particle uniform distribution index of each area respectively; The particle size distribution evaluation and coefficient generation module builds a particle size distribution evaluation model based on the crystal growth driving force coefficient and particle uniform distribution index of each region, generates the particle size distribution coefficient of each region, and performs a comprehensive analysis after the generation to generate the evaluation coefficient; The uniformity evaluation and signal generation module analyzes the generated evaluation coefficients, evaluates the uniformity of the particle size distribution inside the reactor during the crystallization of lithium hexafluorophosphate, and divides the uniformity of the particle size distribution into high uniformity, normal uniformity and non-uniformity according to the evaluation results, and generates high uniformity signals, normal uniformity signals and non-uniform signals respectively; The out-of-control risk warning and control module, when generating uneven signals, continuously obtains several evaluation coefficients generated subsequently, conducts comprehensive analysis, generates different levels of out-of-control risk warning signals for particle size distribution, and takes corresponding control measures for different warning levels; The real-time monitoring and model optimization module continuously monitors the crystallization dynamic information and particle size distribution coefficient of each area inside the reactor, and dynamically adjusts the parameters and control strategies of the particle size distribution evaluation model according to the real-time monitoring results. At the same time, it stores and analyzes historical data to optimize model parameters and control strategies.

Images