Method for analyzing correlation between residual alkali of positive electrode material and moisture

Abstract

The application discloses a method for analyzing the correlation between residual alkali and moisture of a positive electrode material, and relates to the technical field of battery material detection. The method collects the resonance characteristic parameters of a positive electrode material powder sample through non-contact microwave resonance detection, extracts target parameters as input characteristic vectors, obtains residual alkali and moisture content prediction values in combination with a partial least squares regression model, determines the specific contents of lithium hydroxide, lithium carbonate and moisture by using a correlation standard curve, and realizes the correlation analysis of the two. The method can adapt to different dielectric responses, modeling and temperature compensation states, support online continuous detection, stability determination and production linkage, solve the problems of low detection efficiency and insufficient correlation analysis in traditional detection, improve the detection accuracy and production adaptability, and is suitable for large-scale production detection of positive electrode materials.

Description

Technical Field

[0001] This invention relates to the field of battery material testing technology, and more specifically, to a method for analyzing the correlation between residual alkali and moisture in cathode materials. Background Technology

[0002] With the global energy structure transformation and the advancement of the dual-carbon strategy, lithium-ion batteries are increasingly widely used in new energy vehicles, large-scale energy storage, and other fields. As a core component that determines battery performance, safety, and cost, the quality control of cathode materials is crucial. Residual alkali (mainly lithium hydroxide and lithium carbonate) and moisture are key physicochemical indicators of cathode materials, and there is a significant correlation between the two. Residual alkali easily absorbs moisture from the air and reacts, leading to material structure damage, decreased electrochemical performance, and affecting the electrode slurry preparation process. This can exacerbate problems such as battery swelling and shortened cycle life, seriously restricting the quality and safety of battery products.

[0003] Currently, the detection of residual alkali and moisture in cathode materials mainly employs traditional methods such as acid-base titration and Karl Fischer titration. These methods require sample pretreatment, are cumbersome, have long detection cycles, and can only detect residual alkali or moisture content separately, failing to perform correlation analysis and reflect the synergistic effect of the two on cathode material performance. Furthermore, traditional detection methods are mostly offline, unsuitable for the online continuous detection requirements of large-scale production lines, and easily affected by ambient temperature and sample condition, resulting in insufficient accuracy. In addition, existing detection methods lack dynamic optimization of the detection model and temperature compensation mechanisms, making it difficult to adapt to sample detection under different dielectric response states. This hinders the provision of real-time and reliable data support for production process adjustments, thus restricting the improvement of cathode material production efficiency and product consistency. Summary of the Invention

[0004] In order to overcome the problems of inefficient, inaccurate and non-correlation analysis of residual alkali and moisture in cathode materials in existing technologies, this invention discloses a correlation analysis method for residual alkali and moisture in cathode materials, which can effectively solve the above-mentioned technical problems.

[0005] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows: A method for analyzing the correlation between residual alkali and moisture in cathode materials, characterized by comprising: Non-contact microwave resonance detection is performed on positive electrode material powder samples. Based on the collected microwave resonance signals, corresponding resonance characteristic parameters are generated. The resonance characteristic parameters include the resonance frequency shift, quality factor change and reflection coefficient phase angle at different operating frequencies. Obtain the target parameters from the resonant characteristic parameters, which include the resonant frequency offset, quality factor change, reflection coefficient phase angle, and ambient temperature corresponding to each operating frequency; Based on the dielectric response state of the cathode material powder sample, the target parameter is used as the input feature vector, and combined with the partial least squares regression model, the predicted values ​​of residual alkali content and moisture content of the cathode material powder sample are obtained. By correlating each predicted value of residual alkali content and the predicted value of moisture content with the corresponding standard curve, the lithium hydroxide content, lithium carbonate content, and moisture content are obtained. Based on the lithium hydroxide content, lithium carbonate content, and moisture content, a correlation analysis of residual alkali and moisture in the cathode material is achieved.

[0006] Furthermore, when the positive electrode material powder sample is in different dielectric response states, at least one of the following processing steps is also included: When the dielectric constant of water and the dielectric constant of residual alkali satisfy a preset multiple relationship at one of the operating frequencies, a mapping relationship between the input feature vector and the equivalent dielectric constant of the mixed medium is constructed based on the modified Maxwell-Garnett dielectric mixing theory. When the dielectric relaxation characteristic peak of water and the dielectric relaxation characteristic of residual alkali at another operating frequency meet the preset difference condition, based on the Debye relaxation model, the independent contribution of water to the dielectric response is decoupled from the input feature vector to obtain the corrected residual alkali characteristic parameters.

[0007] Furthermore, when the partial least squares regression model is in different modeling states, at least one of the following processing steps is also included: When the correlation coefficients between the input feature vector and the lithium hydroxide content, lithium carbonate content and moisture content meet different preset thresholds, the weight matrix of the partial least squares regression model is constructed using the corresponding preset latent variable extraction strategy. When the correlation coefficient between any two parameters in the input feature vector exceeds a preset threshold, principal component analysis is performed on the input feature vector to reduce its dimensionality, and after eliminating multicollinearity, the weight matrix of the partial least squares regression model is constructed.

[0008] Furthermore, when the cathode material powder sample is under different temperature compensation states, at least one of the following processing steps is also included: A reference resonator with an isolated sample channel is set up in the same cavity of the microwave resonant cavity to monitor the effect of ambient temperature on the resonant frequency drift in real time, and the resonant frequency shift after temperature compensation is calculated. When the ambient temperature exceeds different preset temperature ranges, the corresponding temperature normalization processing is adopted, the partial least squares regression sub-model of the corresponding temperature range is called for prediction, or the heating and dehumidification calibration program is started and the resonance characteristic parameters are re-acquired.

[0009] Furthermore, it also includes the steps for establishing a partial least squares regression model: Acquire automatic calibration data of standard samples based on microwave resonance detection unit, wherein the automatic calibration data of standard samples includes resonance characteristic parameters of cross-grid standard samples with different residual alkali content gradients and different moisture content gradients; Obtain the calibration reference values ​​from the automatic calibration data of the standard sample. The calibration reference values ​​include reference values ​​for lithium hydroxide content and lithium carbonate content measured by acid-base titration, and reference values ​​for moisture content measured by Karl Fischer method. When the residual alkali content gradient and moisture content gradient of the standard sample satisfy the cross-grid design conditions, the resonant characteristic parameters of the standard sample are used as the input matrix, and the calibration reference value is used as the output matrix. The weight matrix and bias vector are solved by partial least squares regression to establish the partial least squares regression model.

[0010] Furthermore, it also includes optimization steps for the partial least squares regression model: When the sum of squared predicted residuals of the partial least squares regression model satisfies the convergence condition, the current weight matrix and bias vector are permanently stored in the system database. When the cross-validation determination coefficient of the partial least squares regression model is lower than the preset threshold, the grid density of the standard samples is increased, the number of gradients for residual alkali content and moisture content is expanded, and partial least squares regression modeling is performed again.

[0011] Furthermore, it also includes a step for determining the stability of the cathode material: The residual alkali-moisture correlated degradation index is calculated based on the lithium hydroxide content, lithium carbonate content, and moisture content. The formula for calculating the correlated degradation index is as follows: D = α ·( C 水 / C 水,ref )+ β ·( C residue / C residue,ref )+ c ·[( C 水 / C 水,ref )·( C residue / C residue,ref )]; in, C 水 Moisture content, C 水,ref This is a reference value for moisture content. C residue This refers to the total residual alkali content.C residue,ref This is a reference value for total residual alkali content. α , β , c Preset weighting coefficients; The stability level of the cathode material powder sample is determined based on the associated degradation index, and a status indicator and related prompts, warnings or alarm signals are output accordingly. The sample is blocked from entering the next production stage based on the alarm signal.

[0012] Furthermore, the determination of the stability level based on the associated degradation index includes: When the associated degradation index is within different preset degradation threshold ranges, different levels of status indicators are output. At the same time, depending on the level, at least one operation in the production process is executed, such as real-time display of parameters, generation of attention prompts, generation of early warning signals and provision of handling suggestions, generation of non-conforming alarm signals and blocking of the production process.

[0013] Furthermore, it also includes production linkage steps: The lithium hydroxide content, lithium carbonate content, and moisture content are output to the production line distributed control system in real time in a preset signal format. The relevant detection parameters and associated degradation indices are stored in the database in time series form to generate historical trend curves. When any detection parameter or associated degradation index exceeds the corresponding preset process control limit, the production line process parameter adjustment command is automatically triggered.

[0014] Furthermore, when the cathode material powder sample is under online continuous monitoring, the following steps are also included: The positive electrode material powder sample is continuously passed through the sample flow channel of the microwave resonant cavity in a preset conveying manner, and the sample flow channel is a non-contact channel. During the continuous passage of the sample through the microwave resonant cavity, microwave signals at different operating frequencies are excited at a preset sampling frequency and resonant characteristic parameters are collected. The continuously acquired resonant characteristic parameters are filtered and Fourier transformed to extract the target parameters at each sampling time as the input feature vector. The input feature vector is substituted into the partial least squares regression model at each time step to calculate and output the time series of lithium hydroxide content, lithium carbonate content and moisture content of the sample in the continuous production process in real time.

[0015] Compared with existing technologies, the beneficial effects of this invention are as follows: This method utilizes non-contact microwave resonance detection technology, eliminating the need for sample pretreatment and enabling rapid acquisition of resonance characteristic parameters. Combined with a partial least squares regression model, it achieves simultaneous prediction of residual alkali and moisture content. This avoids sample contamination and component loss during traditional pretreatment processes, and improves detection accuracy through multi-frequency parameter acquisition, effectively solving the problems of low detection efficiency, poor accuracy, and inability to perform simultaneous detection in traditional methods. By adapting processing strategies to different dielectric response states, and decoupling the dielectric response contributions of moisture and residual alkali based on the modified Maxwell-Garnett dielectric mixing theory and Debye relaxation model, the mutual interference between the two is eliminated, further improving detection accuracy. Furthermore, by combining a temperature compensation mechanism and model optimization steps, the influence of ambient temperature can be offset, and model parameters can be dynamically optimized, ensuring the stability and reliability of detection results under different operating conditions. By calculating the residual alkali-moisture correlation degradation index and determining the stability level, the system can quantify the synergistic effect of the two on the cathode material, and can promptly issue prompts and alarm signals to prevent unqualified samples from flowing into the next stage, thus ensuring product quality. It supports online continuous detection and production linkage, and can output detection data in real time and trigger process adjustment instructions to achieve closed-loop control of the production process, improve production efficiency and product consistency, and provide strong support for the large-scale and high-quality production of cathode materials. Attached Figure Description

[0016] Figure 1 A flowchart of the method for analyzing the correlation between residual alkali and moisture in cathode materials provided in the embodiments of this application. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0018] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0019] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined with each other.

[0020] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0021] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use, or the orientation or positional relationship commonly understood by those skilled in the art. They are only used for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention. In addition, the terms "first," "second," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0022] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0023] This method utilizes non-contact microwave resonance detection technology, eliminating the need for any pretreatment of the cathode material powder sample. It rapidly and non-destructively acquires the resonant characteristic parameters of the sample and, combined with a partial least squares regression model, simultaneously predicts the residual alkali and moisture content. This fundamentally avoids problems such as sample contamination and component loss that occur during traditional pretreatment processes. Furthermore, multi-frequency parameter acquisition improves detection accuracy, effectively addressing the technical pain points of traditional detection methods, such as low efficiency, poor accuracy, and the inability to simultaneously detect residual alkali and moisture. The specific implementation steps are as follows: Please see Figure 1Non-contact microwave resonance detection and resonance characteristic parameter acquisition. Optionally, the entire detection system is first started. This system mainly consists of a microwave signal generator, a microwave resonant cavity, a signal acquisition module, a temperature monitoring module, and a data processing unit. The microwave signal generator can stably output microwave signals at least two different operating frequencies (preferably a dual-frequency combination of 2.45GHz and 5.8GHz, which can be flexibly adjusted according to the sample type). The microwave signal power is controlled between 10-50mW to avoid excessive power causing sample heating and changes in composition, which would affect the accuracy of the detection results. The cathode material powder sample to be tested (including but not limited to ternary cathode materials, lithium iron phosphate cathode materials, and other types of lithium-ion battery cathode material powders) is placed in the non-contact sample channel of the microwave resonant cavity, ensuring that the sample has no direct contact with the inner wall of the microwave resonant cavity, avoiding interference with the transmission and reflection of microwave signals, and ensuring the non-destructive nature of the detection. Subsequently, the microwave signal generator excites the microwave signal at a preset frequency. After the microwave signal passes through the sample, the signal acquisition module collects the reflected microwave resonant signal in real time and transmits the collected raw signal to the data processing unit. After a series of processing steps, including filtering, amplification, and demodulation, the corresponding resonant characteristic parameters are generated. These resonant characteristic parameters specifically include the resonant frequency shift, quality factor change, and reflection coefficient phase angle at different operating frequencies. The resonant frequency shift is the difference between the microwave resonant frequency when the sample is present and the resonant frequency when the microwave resonant cavity is empty (without the sample). The quality factor change is the difference between the quality factor when the microwave resonant cavity is empty and the quality factor when the sample is present. The reflection coefficient phase angle is the phase shift of the reflected microwave signal relative to the incident microwave signal. The acquisition accuracy of all parameters is controlled within ±0.1% to ensure data reliability. Simultaneously, the temperature monitoring module synchronously collects the temperature of the detection environment (i.e., ambient temperature) in real time. T The temperature data is then synchronously transmitted to the data processing unit.

[0024] Target parameter extraction and input feature vector construction. Optionally, the data processing unit filters and extracts the acquired resonant feature parameters to obtain target parameters. These target parameters specifically include the resonant frequency shift, quality factor variation, reflection coefficient phase angle, and synchronously acquired ambient temperature for each operating frequency. T The target parameters are arranged in a preset order to construct a unified input feature vector. X This provides complete and comprehensive input data for subsequent content prediction, ensuring that the prediction model can fully utilize the dielectric response information of the sample and improve prediction accuracy.

[0025] Parameter processing and content prediction based on dielectric response state. Optionally, the actual dielectric response state of the cathode material powder sample is first determined. Based on different dielectric response states, corresponding processing strategies are adopted to eliminate the mutual interference between moisture and residual alkali, further improving detection accuracy. This is because the dielectric response characteristics of moisture and residual alkali are easily superimposed; without targeted processing, the prediction results will have a large deviation. Specifically, when the dielectric constant of water and the dielectric constant of residual alkali at one of the operating frequencies meet a preset multiple relationship (this preset multiple relationship can be calibrated through preliminary experiments according to the sample type and detection accuracy requirements, and is usually set to 3-5 times), an input feature vector is constructed based on the modified Maxwell-Garnett dielectric mixing theory. X The mapping relationship between the dielectric constant of the mixed medium and the dielectric constant of the mixed medium allows for precise characterization of the mixed dielectric properties of water and residual alkali in the sample, reducing mutual interference. When the dielectric relaxation characteristic peaks of water and residual alkali at another operating frequency meet a preset difference condition (i.e., the difference in the position and intensity of their dielectric relaxation characteristic peaks reaches a preset threshold, which can be determined through spectral analysis), based on the Debye relaxation model, the input feature vector... X The independent contribution of moisture to the dielectric response is decoupled to obtain the corrected residual alkali characteristic parameters, thus completely eliminating the interference of moisture on residual alkali detection. The processed input feature vector is substituted into a pre-established partial least squares regression model. Through model calculation, the predicted values ​​of residual alkali content and moisture content of the cathode material powder sample are established. The modeling process of the partial least squares regression model needs to be completed in advance and needs to be optimized according to different modeling states to ensure the accuracy of model prediction.

[0026] The establishment and optimization of the partial least squares regression model. Optionally, the model establishment process is as follows: First, standard samples are prepared. The standard samples adopt a cross-grid design with different residual alkali content gradients and different moisture content gradients to ensure that the standard samples can cover the range of residual alkali and moisture content that may occur in actual testing. Usually, at least 5 residual alkali content gradients and 5 moisture content gradients are set to form a cross-grid standard sample set. Non-contact microwave resonance detection is performed on all cross-grid standard samples through a microwave resonance detection unit to collect the resonance characteristic parameters of each standard sample and form automatic calibration data of the standard samples. At the same time, acid-base titration is used to calibrate each standard sample. Standard samples were tested to obtain reference values ​​for lithium hydroxide (LiOH) content and lithium carbonate (Li2CO3) content. The Karl Fischer method was used to determine the reference value for moisture content in each standard sample. These measured values ​​were used as calibration reference values ​​for model calibration and validation. When the residual alkali content gradient and moisture content gradient of the standard samples met the cross-grid design conditions (i.e., uniform gradient distribution, comprehensive coverage, and no omission of key content ranges), the resonant characteristic parameters of the standard samples were used as the input matrix, and the calibration reference values ​​were used as the output matrix. Partial least squares regression was used to solve the model's weight matrix. W and bias vector b The partial least squares regression model was established.

[0027] The model optimization process is as follows: After establishing the model, the model is validated, and the sum of squared predicted residuals and cross-validation determination coefficients are calculated. R 2 When the sum of squared predicted residuals of the partial least squares regression model meets the convergence condition (i.e., the sum of squared predicted residuals is less than a preset convergence threshold, which is set according to the detection accuracy requirements, usually 10), the convergence condition is met. -4 (The following) will be the current weight matrix. W and bias vector b The data is stored in the system database as a baseline model for subsequent testing; the cross-validation determination coefficient of the partial least squares regression model... R 2 When the value is below a preset threshold (usually set to 0.95), it indicates insufficient model prediction accuracy. In this case, the grid density of the standard samples is increased, expanding the original residual alkali content gradient and moisture content gradient to at least 7. Cross-grid standard samples are prepared, and the resonance characteristic parameters and calibration reference values ​​of the standard samples are re-acquired. Partial least squares regression is then used again for modeling until the model's cross-validation determination coefficient is reached. R 2To ensure the model accurately predicts residual alkali and moisture content, preset thresholds are reached. Furthermore, during model usage, if the correlation coefficients between the input feature vector and the lithium hydroxide, lithium carbonate, and moisture content meet different preset thresholds, the corresponding preset latent variable extraction strategy is used to construct the weight matrix of the partial least squares regression model. If the correlation coefficient between any two parameters in the input feature vector exceeds the preset threshold (usually set to 0.85), it indicates multicollinearity among the parameters, which will affect the model's prediction accuracy. In this case, principal component analysis is performed on the input feature vector to reduce dimensionality and eliminate multicollinearity before constructing the weight matrix of the partial least squares regression model to further optimize model performance.

[0028] Temperature compensation processing. Optionally, since ambient temperature can cause drift in the microwave resonant signal, thus affecting the accuracy of the resonant characteristic parameter acquisition and ultimately leading to deviations in the content prediction results, this method incorporates a temperature compensation mechanism. Based on the different temperature compensation states of the cathode material powder sample, corresponding temperature compensation steps are executed. Specifically, a reference resonator is set within the same cavity of the microwave resonant cavity, isolating the sample channel. The reference resonator does not contact any sample and is only used to monitor the real-time effect of ambient temperature on the resonant frequency drift, using a preset formula. ,in This refers to the resonant frequency shift after temperature compensation. To measure the resonant frequency shift, To calculate the temperature drift monitored by the reference resonator, the temperature-compensated resonant frequency shift is obtained, eliminating the influence of temperature on the resonant frequency shift parameter. When the ambient temperature T exceeds the first temperature range (usually set to 20~25℃, i.e., the conventional detection temperature range), the input feature vector is... X Temperature normalization is performed to incorporate the temperature effect into the parameter correction range, ensuring parameter consistency; when the ambient temperature... T When the temperature exceeds the second temperature range (which is higher than the upper limit of the first temperature range, usually set to 30°C), the heating and dehumidification calibration program is started to heat the positive electrode material powder sample to the preset temperature (usually set to 40~50°C). After the sample temperature stabilizes, the resonant characteristic parameters are reacquired to completely eliminate the interference of high temperature and high humidity environment on the test results and ensure the stability and reliability of the test results under different working conditions.

[0029] Content calibration and correlation analysis between residual alkali and moisture. Optionally, each predicted value of residual alkali content and moisture content obtained through partial least squares regression model is correlated with the corresponding standard curve. The predicted value of residual alkali content is correlated with the standard curves of lithium hydroxide and lithium carbonate, and the predicted value of moisture content is correlated with the standard curve of moisture. Through standard curve conversion, the specific values ​​of lithium hydroxide content, lithium carbonate content, and moisture content in the cathode material powder sample are obtained. Based on the obtained lithium hydroxide content, lithium carbonate content, and moisture content, the correlation between residual alkali and moisture is analyzed. Specifically, this can be done by calculating the correlation coefficient between the two and analyzing their changing trends to clarify the mutual influence relationship between residual alkali and moisture, such as the change law of moisture content when the residual alkali content increases. This provides data support for subsequent sample stability determination and production process adjustment, realizing the correlation analysis between residual alkali and moisture in cathode materials.

[0030] Subsequent processing steps include cathode material stability assessment, production linkage, and online continuous monitoring to achieve a closed loop of testing and control, fully leveraging the technical advantages of this method to ensure product quality and improve production efficiency. Specifically, the cathode material stability assessment step is as follows: Based on the measured lithium hydroxide and lithium carbonate contents, the total residual alkali content is calculated (total residual alkali content = lithium hydroxide content + lithium carbonate content, summed after conversion according to corresponding molar mass). Then, combined with the moisture content, it is substituted into the formula for calculating the residual alkali-moisture related degradation index D. The formula is as follows: D = α ·( C 水 / C 水,ref )+ β ·( C residue / C residue,ref )+ c ·[( C 水 / C 水,ref )·( C residue / C residue,ref )]; in, C 水 Moisture content, C 水,ref This is a reference value for moisture content. C residue This refers to the total residual alkali content. C residue,ref This is a reference value for total residual alkali content. α , β , cThe preset weighting coefficients (set according to the influence of residual alkali and moisture on the stability of the cathode material, with the sum of the three being 1) are used to calculate the associated degradation index. D Afterwards, according to D The stability level of the cathode material powder sample is determined by the preset degradation threshold range it falls within. Specifically, when... D When the first degradation threshold is ≤1 (usually set to 0.8), it is classified as Grade 1 (Excellent); when the first degradation threshold is < D When the value is ≤ the second degradation threshold (usually set to 1.0), it is classified as Grade 2 (Good); when the second degradation threshold is < D When the degradation level is ≤ the third degradation threshold (usually set to 1.2), it is classified as level three (medium). D > When the third degradation threshold is reached, it is classified as level four (poor); according to different stability levels, corresponding status indicators and related prompts, warnings, or alarm signals are output, and corresponding operations are performed simultaneously: at level one, an excellent status indicator is output, and the lithium hydroxide content, lithium carbonate content, moisture content, and associated degradation index are displayed. D The status is displayed in real-time on the operation interface for staff to view. At the second level, a "good" status indicator is output and a attention prompt is generated, reminding staff to continuously monitor the sample status. At the third level, a "medium" status indicator is output and a warning signal is generated, along with processing suggestions, recommending drying or surface modification of the cathode material powder sample to improve sample stability. At the fourth level, a "poor" status indicator is output and a non-conforming alarm signal is generated, simultaneously triggering an interlocking mechanism to prevent the cathode material powder sample from entering the next production stage, thus avoiding the impact of non-conforming samples on the final product quality. This step, by quantifying the synergistic effect of residual alkali and moisture on the cathode material, can promptly issue prompts and alarm signals, effectively preventing non-conforming samples from flowing into the next stage and ensuring product quality.

[0031] The specific production linkage steps are as follows: The measured lithium hydroxide content, lithium carbonate content, and moisture content are output in real time to the production line distributed control system in a preset signal format (preferably 4~20mA analog signal or Modbus TCP / IP digital signal format, adapted to the existing production line distributed control system), so that the production system can acquire sample test data in real time; at the same time, the lithium hydroxide content, lithium carbonate content, moisture content, and associated degradation index are also output. D The data is stored in the system database in time-series format. The database automatically generates historical trend curves for each parameter, facilitating staff to trace and analyze sample quality change patterns and providing data for production process optimization. This data is used to analyze lithium hydroxide content, lithium carbonate content, moisture content, or related degradation indices. DWhen any of the parameters exceeds the corresponding preset process control limit, the system automatically triggers a process parameter adjustment command for the production line. The command is sent to the corresponding execution unit on the production line to adjust the relevant production process parameters (such as drying temperature, reaction time, etc.), thereby achieving real-time control of the production process and forming a closed loop of detection-control-optimization. This step achieves closed-loop control of the production process by outputting detection data in real time and triggering process adjustment commands, effectively improving production efficiency and product consistency.

[0032] The online continuous detection steps are as follows: When the cathode material powder sample is in an online continuous production scenario, the cathode material powder sample is continuously transported through the sample flow channel of the microwave resonant cavity using a preset transport method (preferably free fall or pneumatic transport to ensure continuous and uniform sample transport). This sample flow channel is a fluidized bed type non-contact channel to avoid contamination or signal interference caused by sample contact with the channel. During the continuous passage of the sample through the microwave resonant cavity, the microwave signal generator alternately or simultaneously excites microwave signals of different operating frequencies at a preset sampling frequency (usually set to 10~20Hz, which can be adjusted according to the production speed). The signal acquisition module continuously acquires the microwave resonant signal at each moment. The data processing unit performs sliding window filtering (to eliminate signal noise) and fast Fourier transform processing (to extract effective signal features) on the continuously acquired resonant characteristic parameters, extracting the target parameters at each sampling moment (resonant frequency shift, quality factor change, reflection coefficient phase angle, and ambient temperature corresponding to each operating frequency), and organizing them into an input feature vector. X t ; Input feature vector X t The partial least squares regression model is substituted into the samples at each time step to calculate and output the time series of LiOH content, Li2CO3 content, and moisture content of the cathode material powder sample in the continuous production process in real time. This allows staff to monitor the sample quality in the continuous production process in real time and detect production abnormalities in a timely manner.

[0033] In summary, this implementation method fully discloses all the steps of the method, from sample testing, parameter processing, model establishment and optimization, to temperature compensation, content calibration, correlation analysis and subsequent stability determination, production linkage, and online continuous detection. Furthermore, this method eliminates the need for sample pretreatment through non-contact microwave resonant detection, avoiding sample contamination and component loss caused by traditional pretreatment. Combined with multi-frequency parameter acquisition, it significantly improves detection accuracy. Dielectric response decoupling eliminates the mutual interference between moisture and residual alkali, and temperature compensation and model optimization ensure stable and reliable results under different operating conditions. By quantifying the synergistic effects through the correlation degradation index, it achieves graded early warning and interception of non-conforming products. Combined with online detection and production linkage, it forms a closed-loop management system, improving the production efficiency and product consistency of cathode materials, and providing technical support for large-scale, high-quality production.

[0034] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for analyzing the correlation between residual alkali and moisture in cathode materials, characterized in that, include: Non-contact microwave resonance detection is performed on positive electrode material powder samples. Based on the collected microwave resonance signals, corresponding resonance characteristic parameters are generated. The resonance characteristic parameters include the resonance frequency shift, quality factor change and reflection coefficient phase angle at different operating frequencies. Obtain the target parameters from the resonant characteristic parameters, which include the resonant frequency offset, quality factor change, reflection coefficient phase angle, and ambient temperature corresponding to each operating frequency; Based on the dielectric response state of the cathode material powder sample, the target parameter is used as the input feature vector, and combined with the partial least squares regression model, the predicted values ​​of residual alkali content and moisture content of the cathode material powder sample are obtained. By correlating each predicted value of residual alkali content and the predicted value of moisture content with the corresponding standard curve, the lithium hydroxide content, lithium carbonate content, and moisture content are obtained. Based on the lithium hydroxide content, lithium carbonate content, and moisture content, a correlation analysis of residual alkali and moisture in the cathode material is achieved.

2. The method for analyzing the correlation between residual alkali and moisture in cathode materials according to claim 1, characterized in that, When the positive electrode material powder sample is in different dielectric response states, at least one of the following processing steps is also included: When the dielectric constant of water and the dielectric constant of residual alkali satisfy a preset multiple relationship at one of the operating frequencies, a mapping relationship between the input feature vector and the equivalent dielectric constant of the mixed medium is constructed based on the modified Maxwell-Garnett dielectric mixing theory. When the dielectric relaxation characteristic peak of water and the dielectric relaxation characteristic of residual alkali at another operating frequency meet the preset difference condition, based on the Debye relaxation model, the independent contribution of water to the dielectric response is decoupled from the input feature vector to obtain the corrected residual alkali characteristic parameters.

3. The method for analyzing the correlation between residual alkali and moisture in cathode materials according to claim 1, characterized in that, When the partial least squares regression model is in different modeling states, at least one of the following processing steps is also included: When the correlation coefficients between the input feature vector and the lithium hydroxide content, lithium carbonate content and moisture content meet different preset thresholds, the weight matrix of the partial least squares regression model is constructed using the corresponding preset latent variable extraction strategy. When the correlation coefficient between any two parameters in the input feature vector exceeds a preset threshold, principal component analysis is performed on the input feature vector to reduce its dimensionality, and after eliminating multicollinearity, the weight matrix of the partial least squares regression model is constructed.

4. The method for analyzing the correlation between residual alkali and moisture in cathode materials according to claim 1, characterized in that, When the cathode material powder sample is under different temperature compensation states, at least one of the following processing steps is also included: A reference resonator with an isolated sample channel is set up in the same cavity of the microwave resonant cavity to monitor the effect of ambient temperature on the resonant frequency drift in real time, and the resonant frequency shift after temperature compensation is calculated. When the ambient temperature exceeds different preset temperature ranges, the corresponding temperature normalization processing is adopted, the partial least squares regression sub-model of the corresponding temperature range is called for prediction, or the heating and dehumidification calibration program is started and the resonance characteristic parameters are re-acquired.

5. The method for analyzing the correlation between residual alkali and moisture in cathode materials according to claim 1, characterized in that, It also includes the steps for establishing a partial least squares regression model: Acquire automatic calibration data of standard samples based on microwave resonance detection unit, wherein the automatic calibration data of standard samples includes resonance characteristic parameters of cross-grid standard samples with different residual alkali content gradients and different moisture content gradients; Obtain the calibration reference values ​​from the automatic calibration data of the standard sample. The calibration reference values ​​include reference values ​​for lithium hydroxide content and lithium carbonate content measured by acid-base titration, and reference values ​​for moisture content measured by Karl Fischer method. When the residual alkali content gradient and moisture content gradient of the standard sample satisfy the cross-grid design conditions, the resonant characteristic parameters of the standard sample are used as the input matrix, and the calibration reference value is used as the output matrix. The weight matrix and bias vector are solved by partial least squares regression to establish the partial least squares regression model.

6. The method for analyzing the correlation between residual alkali and moisture in cathode materials according to claim 5, characterized in that, It also includes the optimization steps for the partial least squares regression model: When the sum of squared predicted residuals of the partial least squares regression model satisfies the convergence condition, the current weight matrix and bias vector are permanently stored in the system database. When the cross-validation determination coefficient of the partial least squares regression model is lower than the preset threshold, the grid density of the standard samples is increased, the number of gradients for residual alkali content and moisture content is expanded, and partial least squares regression modeling is performed again.

7. The method for analyzing the correlation between residual alkali and moisture in cathode materials according to any one of claims 1 to 6, characterized in that, It also includes steps for determining the stability of the cathode material: The residual alkali-moisture correlated degradation index is calculated based on the lithium hydroxide content, lithium carbonate content, and moisture content. The formula for calculating the correlated degradation index is as follows: D = α ·( C 水 / C 水,ref )+ β ·( C residue / C residue,ref )+ γ ·[( C 水 / C 水,ref )·( C residue / C residue,ref )]； in, C 水 Moisture content, C 水,ref This is a reference value for moisture content. C residue This refers to the total residual alkali content. C residue,ref This is a reference value for total residual alkali content. α , β , γ Preset weighting coefficients; The stability level of the cathode material powder sample is determined based on the associated degradation index, and a status indicator and related prompts, warnings or alarm signals are output accordingly. The sample is blocked from entering the next production stage based on the alarm signal.

8. The method for analyzing the correlation between residual alkali and moisture in cathode materials according to claim 7, characterized in that, The determination of stability level based on the associated degradation index includes: When the associated degradation index is within different preset degradation threshold ranges, different levels of status indicators are output. At the same time, depending on the level, at least one operation in the production process is executed, such as real-time display of parameters, generation of attention prompts, generation of early warning signals and provision of handling suggestions, generation of non-conforming alarm signals and blocking of the production process.

9. The method for analyzing the correlation between residual alkali and moisture in cathode materials according to any one of claims 1 to 6, characterized in that, It also includes production linkage steps: The lithium hydroxide content, lithium carbonate content, and moisture content are output to the production line distributed control system in real time in a preset signal format. The relevant detection parameters and associated degradation indices are stored in the database in time series form to generate historical trend curves. When any detection parameter or associated degradation index exceeds the corresponding preset process control limit, the production line process parameter adjustment command is automatically triggered.

10. The method for analyzing the correlation between residual alkali and moisture in cathode materials according to claim 1, characterized in that, When the cathode material powder sample is under online continuous monitoring, the following steps are also included: The positive electrode material powder sample is continuously passed through the sample flow channel of the microwave resonant cavity in a preset conveying manner, and the sample flow channel is a non-contact channel. During the continuous passage of the sample through the microwave resonant cavity, microwave signals at different operating frequencies are excited at a preset sampling frequency and resonant characteristic parameters are collected. The continuously acquired resonant characteristic parameters are filtered and Fourier transformed to extract the target parameters at each sampling time as the input feature vector. The input feature vector is substituted into the partial least squares regression model at each time step to calculate and output the time series of lithium hydroxide content, lithium carbonate content and moisture content of the sample in the continuous production process in real time.

Images