Multi-parameter regulation and control intelligent particle size detection method and system

Abstract

The invention discloses a multi-parameter regulation and control intelligent particle size detection method and system, and belongs to the technical field of particle size measurement. The method comprises the following steps: acquiring to-be-detected particle material features to match initial dispersion parameters, and dynamically updating the parameters in combination with image recognition and real-time shading degree data to obtain a stably dispersed suspension; in an optical measurement area, adjusting the light intensity of a laser and the flow rate of a sample based on the shading degree, adjusting temperature control power by combining temperature data, and constructing stable measurement conditions; scattered light signals are collected by using a double-detector group, and effective signals are output through dynamic gain adjustment, self-calibration and cross validation; and finally, effective scattering signals, monitoring data and process parameters are fused, and particle size distribution is solved through an inversion algorithm. According to the invention, full-link closed-loop intelligent regulation and control from sample pretreatment to result generation in the detection process is realized, so that system parameters can be adaptively optimized in real time along with the sample state and environment change, and a stable and accurate particle size distribution result can be obtained in a complex application scene.

Description

Technical Field

[0001] This invention relates to the field of particle size measurement technology, and more specifically to an intelligent particle size detection method and system with multi-parameter control. Background Technology

[0002] Currently, particle size and distribution are key parameters determining the performance of powder materials, and have a significant impact on fields such as cement, catalysts, coatings, pharmaceuticals, and minerals. Light scattering, due to its advantages of non-contact measurement, non-interference with the state of the measured object, short measurement time, and real-time detection, has become the mainstream method for particle size detection. Its core theoretical basis is Mie scattering theory.

[0003] However, existing particle size detection technologies based on light scattering still have limitations: the sample dispersion process relies on human experience, and parameters such as dispersant selection, ultrasonic power, and stirring speed lack an adaptive matching mechanism with the sample material, and there are no effective real-time feedback adjustment methods, which can easily lead to particle agglomeration or insufficient dispersion; the optical measurement parameters are fixed and cannot be dynamically optimized according to the sample's shading degree, concentration, and other conditions, while changes in ambient temperature can easily affect the performance of optical components, resulting in unstable scattering signal quality; signal acquisition uses a single detector or a fixed array of detectors, lacking targeted gain adjustment and interference removal methods, making it difficult to cover the scattering signal requirements of particles of different sizes, and easily affected by abnormal signals; the inversion algorithm in the data processing stage suffers from insufficient accuracy, weak noise resistance, and poor stability, and the inversion calculation relies solely on the scattering signal, which is disconnected from the parameters of the previous dispersion and optical processes, making it difficult to guarantee the accuracy of the detection results.

[0004] Therefore, proposing a multi-parameter-controlled intelligent particle size detection method and system to improve detection accuracy and intelligence level is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, the present invention provides an intelligent particle size detection method and system with multi-parameter control. By constructing a closed-loop adaptive control system covering the entire process from sample dispersion and optical measurement to signal processing and inversion, the method achieves real-time matching and dynamic optimization of detection process parameters and sample state, enabling automatic particle size detection with high consistency and high accuracy under complex working conditions.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: On one hand, this invention discloses a multi-parameter controlled intelligent particle size detection method, comprising the following steps: The material characteristics of the particle sample to be tested are collected to match the initial dispersion parameters; the particle sample to be tested is mixed with the dispersion medium to form an initial suspension; based on the image recognition results of the initial suspension and the real-time shading monitoring data, the initial dispersion parameters are dynamically updated to obtain the suspension to be tested in a preset stable dispersion state. A suspension sample in a stable dispersed state is introduced into the optical measurement area. Based on real-time shading monitoring data, the intensity of the laser irradiating the suspension sample and the flow rate of the sample are adjusted. At the same time, the temperature control power is adjusted based on the temperature monitoring data of the optical measurement area to form stable optical measurement conditions. Under stable optical measurement conditions, the light signal scattered by the sample is collected using a dual detector group. The detector gain is dynamically adjusted according to the real-time signal-to-noise ratio of the light signal. Detector self-calibration and multi-channel signal cross-verification are performed before and after the acquisition, and an effective scattering signal is output. Based on the effective scattering signal, real-time shading monitoring data, temperature monitoring data, and updated dispersion parameters, the particle size distribution of the sample under test is solved by an inversion algorithm.

[0007] Preferably, the initial dispersion parameters for matching the material characteristics of the particle sample to be tested include: The reflectance spectrum characteristics, density, and surface charge properties of the particle sample to be tested are collected to construct a three-dimensional material feature vector; The three-dimensional material feature vector is matched with a pre-stored mapping database to output the corresponding initial dispersion parameter combination, which includes stirring speed, ultrasonic power, dispersant type and dosage.

[0008] Preferably, based on image recognition results and real-time shading monitoring data of the initial suspension, the initial dispersion parameters are dynamically updated to obtain a test suspension in a preset stable dispersion state, including: Extract the equivalent diameter of aggregates, the proportion of aggregates, and the particle uniformity from the image of the initial suspension; The real-time shading degree and shading degree change rate of the initial suspension were collected in conjunction with the data. Using the proportion of agglomerates and the degree of shading as feedback variables, a PID-PWM hybrid algorithm is adopted to gradient adjust the ultrasonic power, stirring speed and dispersant dosage according to preset priorities until the suspension state simultaneously meets the preset conditions of agglomerate proportion, particle uniformity, shading degree and shading degree change rate.

[0009] Preferably, based on real-time shading monitoring data, adjusting the laser light intensity irradiating the suspension sample and the sample flow rate includes: When the real-time shading degree is higher than the preset upper limit and the rate of change of shading degree is non-negative, the laser intensity is controlled to decrease exponentially; when the real-time shading degree is lower than the preset lower limit and the rate of change of shading degree is non-positive, the laser intensity is controlled to increase exponentially. Meanwhile, the sample flow rate is controlled to be linearly adjusted in proportion to the real-time shading value.

[0010] Preferably, the temperature control power is adjusted based on monitoring data of the temperature in the optical measurement area to form stable optical measurement conditions, including: Real-time acquisition of temperature and temperature change rate in the optical measurement area; When the temperature deviates from the set constant temperature range by less than or equal to the first threshold, the temperature control module operates in low-power pulse mode; when the temperature deviates from the first threshold and the temperature change rate is greater than or equal to the second threshold, the temperature control power is adjusted proportionally to the product of the temperature deviation and the temperature change rate.

[0011] Preferably, a dual-detector array is used to acquire the light signal scattered by the sample, and the detector gain is dynamically adjusted based on the real-time signal-to-noise ratio of the light signal, including: Real-time calculation of the signal-to-noise ratio of the signals acquired by the dual detector groups; When the signal-to-noise ratio (SNR) is below the preset lower limit, the control gain is adjusted increasing according to the logarithmic function of the ratio of SNR to the preset lower limit; when the SNR is above the preset upper limit, the control gain is adjusted decreasing according to the logarithmic function of the ratio of SNR to the preset upper limit. The dual detector array includes a central detector array for collecting small-angle scattered light and an edge detector array for collecting large-angle scattered light.

[0012] Preferably, detector self-calibration and multi-channel signal cross-verification are performed before and after acquisition to output effective scattering signals, including: Before each signal acquisition, the dual detector group is illuminated with a standard light source to generate calibration coefficients; After each signal acquisition, the signals acquired by the dual detector groups are time-synchronized and correlation-analyzed to remove abnormal signals with correlation coefficients lower than a preset threshold, and the retained valid signals are corrected using calibration coefficients.

[0013] Preferably, based on the effective scattering signal, real-time shading monitoring data, temperature monitoring data, and updated dispersion parameters, the particle size distribution of the sample under test is solved using an inversion algorithm, including: The updated dispersion parameters, adjusted laser intensity and sample flow rate, adjusted temperature control power, and dynamically adjusted detector gain were collected as process parameters. The analytic hierarchy process was used to assign adaptive weights to each process parameter by combining real-time shading monitoring data and temperature monitoring data. The weighted process parameters and the effective scattering signal are input together into the inversion algorithm for iterative solution, and the particle size distribution of the sample to be tested is output.

[0014] Preferably, the inversion algorithm is an improved bee colony artificial algorithm, including: The search step size is dynamically adjusted based on updated dispersion parameters and optical measurement parameters; By integrating real-time shading and temperature monitoring data, a fitness function with environmental parameter correction terms is constructed. The division of labor among bee colonies is dynamically adjusted based on unimodal, bimodal, and trimodal distributions. The weighted process parameters are incorporated into the iterative optimization until the termination condition is met, and the granularity distribution solution is output.

[0015] On the other hand, the present invention also discloses a multi-parameter controlled intelligent particle size detection system, comprising: The material characteristics and dispersion module is used to collect the material characteristics of the particles to be tested and match the initial dispersion parameters, mix them to form an initial suspension, dynamically update the parameters based on image recognition results and real-time shading data, and output a stable dispersion suspension. The optical measurement and control module is used to receive a stable dispersed suspension and introduce it into the optical measurement area. It adjusts the laser light intensity and sample flow rate based on the real-time shading degree and adjusts the temperature control power in combination with temperature data to form stable optical measurement conditions. The signal acquisition and processing module includes a dual detector group for acquiring scattered light signals, dynamically adjusting the gain according to the signal-to-noise ratio, performing self-calibration and cross-validation before and after acquisition, and outputting effective scattered signals. The data inversion module is used to collect effective scattering signals, monitoring data, and updated dispersion parameters, and solves the granularity distribution through an inversion algorithm.

[0016] As can be seen from the above technical solution, compared with the prior art, this invention discloses a multi-parameter controlled intelligent particle size detection method and system. It adaptively matches and dynamically optimizes dispersion parameters by collecting the material characteristics of the particles to be tested, combines real-time monitoring data to control optical measurement conditions, uses a dual-detector group to dynamically adjust the gain, and obtains effective scattering signals through self-calibration and cross-validation. Finally, it fuses multi-dimensional process parameters and effective signals to solve for particle size distribution through an adaptive inversion algorithm. This invention achieves intelligent collaborative control of the entire process from sample dispersion, optical measurement, signal acquisition to data inversion, which can improve the automation level and environmental adaptability of the detection process, while improving the consistency and accuracy of the detection results. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0018] Figure 1 A flowchart of the method provided by the present invention; Figure 2 This is a schematic diagram of the optical measurement area provided by the present invention; Figure 3 The system architecture diagram provided for this invention. Detailed Implementation

[0019] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] On the one hand, refer to Figure 1 This invention discloses a multi-parameter controlled intelligent particle size detection method, comprising the following steps: S1. Sample dispersion parameters are adaptively adjusted, as follows: S11. Collect material characteristics of the particle sample to be tested to match the initial dispersion parameters, including: S111. Collect the reflection spectrum characteristics, density, and surface charge properties of the particle sample to be tested, and construct a three-dimensional material feature vector.

[0021] When collecting the reflectance spectral characteristics of the particle samples to be tested, an ultraviolet-visible-near-infrared spectrometer is used to scan the sample spectrally. The collected raw spectral data is baseline corrected and smoothed to eliminate interference from ambient light and instrument noise, and key characterization parameters of characteristic absorption peaks in the spectral curve are extracted.

[0022] The density of the particle sample is measured using the buoyancy method. First, the mass of the dried sample is weighed. An inert liquid that does not physically dissolve or chemically react with the sample is selected as the immersion medium. After completely immersing the sample, the volume of the displaced medium is recorded. The density is then calculated using the density formula. Calculate the sample density, where For the density of the particulate sample, This refers to the mass of the sample in its dry state. The volume of the immersion medium displaced from the sample.

[0023] To characterize the surface charge properties of the particle samples, an electrophoretic light scattering instrument was used. The sample was dispersed in a specific electrolyte solution to prepare a dilute suspension, placed in an electrophoretic measurement cell, and the electrophoretic mobility of the particles was recorded after applying a stable electric field. According to Henry's equation ; Calculate the Zeta potential of the particles ,in The relative permittivity of the dispersion medium, The vacuum permittivity, For Henry functions, To disperse the viscosity of the medium.

[0024] When constructing the 3D material feature vector, the reflectance spectrum feature parameters, density, and Zeta potential are normalized to eliminate the differences in dimensions of different physical quantities. A linear normalization method is used, and the calculation formula is as follows: ; in These are the normalized eigenvalues. These are the original eigenvalues. This is the minimum value of this feature dimension. This represents the maximum value of this feature dimension. The three normalized parameters are then used as components of the three-dimensional vector to form the three-dimensional material feature vector. ,in The corresponding normalized value of the reflectance spectral characteristics, Corresponding density normalized value, The corresponding normalized value of the Zeta potential.

[0025] The pre-stored mapping database is constructed using a large amount of experimental data. Different types of common particulate materials are selected, and multiple batches of samples are selected for each material. After obtaining the three-dimensional material feature vectors of each sample, the corresponding optimal dispersion parameter combination is determined through orthogonal experiments. The feature vectors are associated with the optimal dispersion parameter combination and stored one by one to form a relational mapping database that supports fast retrieval and matching.

[0026] Feature vector matching uses Euclidean distance as the similarity evaluation index to calculate the 3D material feature vector X of the sample to be detected and the feature vectors of each sample in the database. Euclidean distance The calculation formula is: ; in , , For the first in the database The feature vector of each sample has three components. The database is traversed to select the dispersion parameter combination corresponding to the sample with the smallest Euclidean distance, which is used as the initial dispersion parameter combination for the sample to be tested, including stirring speed, ultrasonic power, dispersant type and dosage.

[0027] S12. Mix the particle sample to be tested with the dispersion medium to form an initial suspension.

[0028] When selecting a dispersion medium, it must meet the requirements of no physical dissolution, no chemical reaction, and no flocculation or agglomeration with the sample. Distilled water is preferred as the basic dispersion medium. When the sample has poor dispersibility in water, a suitable organic solvent or a specific electrolyte solution can be used. The dispersion medium must be purified before use to remove impurities and air bubbles, ensuring its cleanliness.

[0029] When determining the sample size of the particles to be tested, the shading requirements for subsequent optical measurements are considered. The sample size is calculated based on the volume of the dispersion medium and the preset initial suspension concentration, using the formula... ; Calculation, where For the sample size, To disperse the volume of the medium, The initial suspension mass concentration. To determine the sample density, a high-precision electronic balance was used for weighing during the sampling process to ensure the accuracy of the sample quantity.

[0030] Slowly add the weighed particle sample to a sealed mixing container containing a pre-defined volume of dispersion medium, avoiding evaporation of the dispersion medium or the introduction of external impurities. Start the stirring device and perform initial mixing at the stirring speed specified in the initial dispersion parameter combination, maintaining a steady flow of liquid within the container to prevent the formation of bubbles. After stirring, a uniform and stable initial suspension is formed, ensuring that there are no visible particle agglomerates or sediments in the suspension.

[0031] S13. Based on the image recognition results of the initial suspension and real-time shading monitoring data, dynamically update the initial dispersion parameters to obtain the test suspension in a preset stable dispersion state, including: S131. Extract the equivalent diameter of agglomerates, the proportion of agglomerates, and the particle uniformity from the image of the initial suspension.

[0032] Industrial cameras were used to continuously acquire images of the initial suspension. During acquisition, stable and uniform lighting conditions were ensured to avoid light reflection or shadows affecting image quality. The acquired raw images were preprocessed, with image denoising algorithms used to eliminate noise interference, and image enhancement techniques employed to improve the contrast between particles and the background, laying the foundation for feature extraction.

[0033] When extracting the equivalent diameter of aggregates, the equivalent circle diameter method is used. The outline of each aggregate in the image is fitted to a circle of equal area; the diameter of this circle is the equivalent diameter of the aggregate. When calculating the proportion of aggregates, the number or area of ​​aggregates in the image is counted as a percentage of the total number or area of ​​particles. When characterizing particle uniformity, the coefficient of variation is used as the evaluation index, and the calculation formula is as follows: ; in For particle uniformity, The standard deviation of the equivalent diameter of all particles. The average equivalent diameter of all particles is used to quantify the concentration of particle size distribution.

[0034] S132. Real-time shading degree and shading degree change rate of the initial suspension are collected in conjunction with the system.

[0035] A transmissive shading meter was used to monitor the shading intensity of the initial suspension in real time. The probe beam of the shading meter passed through the suspension sample, and the shading intensity was calculated based on the ratio of transmitted light intensity to incident light intensity. The monitoring process was synchronized with image acquisition to ensure a one-to-one correspondence between the shading intensity data and image data over time.

[0036] When calculating the rate of change of shading, shading data from two consecutive monitoring times are selected, and the formula is used. ; Calculation, where The rate of change of shading. , These are the shading values ​​at two different times, before and after. , These correspond to the respective monitoring times. The rate of change in shading reflects the dynamic trend of suspension concentration in real time, providing dynamic feedback for parameter adjustment.

[0037] S133. Using the proportion of agglomerates and the degree of shading as feedback variables, a PID-PWM hybrid algorithm is adopted to gradient adjust the ultrasonic power, stirring speed and dispersant dosage according to preset priorities until the suspension state simultaneously meets the preset conditions of agglomerate proportion, particle uniformity, shading degree and shading degree change rate.

[0038] The preset parameter adjustment priority is to adjust the ultrasonic power, stirring speed and dispersant dosage in a gradient manner. The ultrasonic power is used first to break up agglomerates, then the stirring speed is used to optimize the particle dispersion uniformity, and finally the dispersant dosage is used to supplement and stabilize the dispersion effect.

[0039] In the PID-PWM hybrid algorithm, the PID algorithm is used to calculate the parameter adjustment, and the core formula is: ; in The PID output adjustment value. This is the proportionality coefficient. The integral coefficient is... These are the differential coefficients. The deviation between the feedback variable and the preset target value is considered. The PWM algorithm converts the adjustment amount output by the PID into a pulse width modulation signal, realizing continuous gradient adjustment of ultrasonic power and stirring speed, as well as quantitative gradient addition of dispersant dosage.

[0040] The system uses the proportion of aggregates and occlusion as the core feedback variables, combined with particle uniformity and the rate of change of occlusion for comprehensive judgment. After each adjustment, images and occlusion data of the suspension are collected simultaneously, and all evaluation indicators are recalculated until all indicators meet the preset conditions. At this point, parameter adjustment is stopped, and the suspension is in a preset stable dispersion state.

[0041] S2. Adaptive optimization of optical measurement parameters, as detailed below: S21. Introduce the suspension sample, which is in a stable dispersed state, into the optical measurement area (e.g., Figure 2 As shown in the figure, the laser intensity irradiating the suspension sample and the sample flow rate are adjusted based on real-time shading monitoring data.

[0042] When introducing a stable, dispersed suspension sample into the optical measurement area, a precision peristaltic pump is used as the power source to ensure a smooth and continuous sample transfer process, avoiding bubbles or particle sedimentation caused by sudden changes in flow rate. The transfer tubing is made of an inert material that does not interact with the sample or dispersion medium, and the tubing connections are designed with a sealed structure to prevent sample leakage or the entry of external contaminants, ensuring the cleanliness of the measurement environment.

[0043] Real-time shading monitoring continues with a transmission-type shading meter, with the monitoring frequency synchronized with the optical measurement cycle to ensure that the shading data reflects the concentration status of the suspension within the optical measurement area in real time. The preset upper and lower limits of shading are determined based on the optimal signal intensity range of the optical measurement, meeting the measurement needs of particles with different size ranges.

[0044] From a physical perspective, the absorption and scattering of light by a suspension follows the Lambert-Beer law, and the transmitted light intensity has an exponential relationship with the concentration. As the core indicator reflecting concentration, the light intensity compensation adjustment uses an exponential function, which precisely matches this physical law, achieving dynamic adaptation between light intensity and concentration. In terms of adjustment effect, exponential adjustment achieves a smooth transition in light intensity, avoiding sudden changes in light intensity that may occur with linear adjustment, and preventing saturation or distortion of the scattered signal due to abrupt changes in light intensity. Simultaneously, the exponential function has a fast initial adjustment rate and a gradual flattening later, enabling rapid response to large changes in shading and slow convergence as it approaches the target light intensity, avoiding signal fluctuations caused by overshoot and ensuring the stability and continuity of the scattered signal. Therefore, this embodiment uses an exponential function control strategy for laser light intensity adjustment. The specific adjustment strategy is as follows: When the real-time shading level is higher than the preset upper limit and the rate of change of shading level is non-negative, the light intensity is adjusted according to the exponential decay law: ; In the formula The adjusted laser light intensity, The initial value of light intensity. The attenuation coefficient is... To adjust the time.

[0045] When the real-time shading level is lower than the preset lower limit and the rate of change of shading level is not positive, the light intensity is adjusted according to an exponential increasing law: .

[0046] Meanwhile, the sample flow rate is linearly adjusted in proportion to the real-time shading value.

[0047] The light-blocking degree is linearly positively correlated with the suspension concentration, while the sample flow rate directly determines the sample renewal rate within the optical measurement area. This linear correspondence between concentration and flow rate changes allows proportional linear adjustment to directly offset the effects of concentration fluctuations, achieving precise control by adjusting the flow rate according to concentration levels, quickly maintaining the concentration in the measurement area within a suitable range. Furthermore, the linear adjustment principle is simple and the response is direct, requiring no complex nonlinear calculations, reducing the computational load and calibration difficulty of the control system. The adjustment process is also highly predictable and easily controlled using the flow rate proportional coefficient. The calibration achieves universal adaptability to different sample types; simultaneously, linear adjustment ensures that changes in flow rate and concentration respond synchronously, avoiding continuous concentration deviations caused by adjustment lag, and providing a guarantee for stable optical measurement conditions. Its adjustment formula is: ; in The adjusted sample flow rate, This represents the real-time shading value. (Scale factor) The calibration is determined based on parameters such as the volume of the optical measurement area and the laser beam path length.

[0048] S22. Simultaneously, based on monitoring data of the temperature in the optical measurement area, the temperature control power is adjusted to form stable optical measurement conditions, including: Temperature acquisition in the optical measurement area is achieved using a high-precision temperature sensor. The sensor, either a contact or non-contact type, is selected for its fast response and high measurement accuracy and is installed in a critical area around the laser beam propagation path. The installation location must avoid direct influence from the sample flow channel while being close to the mounting area of ​​optical components (such as lenses and detectors) to ensure accurate capture of the ambient temperature of the measurement area and to reflect the actual operating temperature of the optical components. The temperature acquisition frequency is synchronized with laser intensity adjustment and sample flow rate regulation to achieve multi-parameter coordinated response and avoid inaccurate control due to temperature data lag.

[0049] The rate of temperature change is calculated using temperature data from two consecutive acquisition cycles. The calculation formula is as follows: ; in For the rate of temperature change, The temperature at the current moment. The temperature at the previous data collection time. This is the current data collection time. This is the previous data collection point. By calculating the rate of temperature change in real time, the trend and severity of temperature changes can be accurately determined, providing data support for switching adjustment strategies.

[0050] The constant temperature range is set based on the optimal operating temperature range of the optical components and the temperature sensitivity of Mie scattering theory. This ensures stable optical component performance while minimizing the impact of temperature on the propagation of scattered light. Both the first and second thresholds are dynamic, not fixed values, and their determination requires dynamic calibration considering the measurement scenario, optical component parameters, and measurement accuracy requirements.

[0051] The determination of the first threshold is based on the temperature stability requirements of the optical components. The initial value is set based on the temperature fluctuation range provided by the optical component manufacturer, and is dynamically adjusted according to the temperature adjustment effect during the actual measurement process. When small temperature deviations occur repeatedly but overshoot occurs frequently after adjustment, the first threshold is appropriately reduced; when the ambient temperature itself has small fluctuations that do not affect the measurement results, the first threshold is appropriately increased to ensure that the adjustment strategy is mild and effective when there is a slight temperature deviation.

[0052] The second threshold is determined based on the risk level of temperature changes, with the initial value set by referencing the temperature change rate tolerance limit of the optical element and the sensitivity of the scattered signal to sudden temperature changes. During measurement, if the temperature change rate exceeds the initial second threshold but does not significantly affect the measurement results, the second threshold can be appropriately increased; if scattering signal distortion occurs when the temperature change rate approaches the initial second threshold, the second threshold is decreased to trigger strong adjustment in advance and avoid the adverse effects of sudden temperature changes. The dynamic adjustment of both thresholds is achieved through an adaptive algorithm in the system's backend, requiring no manual intervention.

[0053] When the temperature deviates from the set constant temperature range by less than or equal to the first threshold, a low-power pulse mode is used. This mode is suitable for scenarios with small temperature fluctuations, achieving minor temperature correction through periodic low-power output, avoiding temperature overshoot or frequent fluctuations caused by continuous high-power regulation. The pulse period and duty cycle are determined based on the power characteristics of the temperature control module and the heat capacity calibration of the measurement area. The period is usually set to be consistent with the temperature acquisition period, while the duty cycle is dynamically adjusted according to the direction (too high or too low) and magnitude of temperature deviation, ensuring regulation accuracy while reducing energy consumption.

[0054] When the temperature deviation exceeds the first threshold and the rate of temperature change is greater than or equal to the second threshold, the temperature control power is adjusted proportionally to the product of the temperature deviation and the rate of temperature change. The calculation formula is as follows: ; in The adjusted temperature control power, This is the temperature control proportional coefficient. This refers to the temperature deviation (i.e., the difference between the actual temperature and the set constant temperature). , This represents the rate of temperature change. Proportional coefficient. Based on the maximum power, the thermal response characteristics of the measurement area, and the temperature protection threshold of the optical components, the temperature control power is calibrated to ensure that it can quickly suppress drastic temperature changes without damaging the optical components due to excessive power.

[0055] The core advantage of this proportional adjustment strategy lies in the fact that the temperature control power adapts synchronously to the temperature deviation and rate of change. The greater the temperature deviation and the more drastic the change, the stronger the adjustment power, which can quickly pull the temperature of the measurement area back to the set range. When the temperature deviation or rate of change decreases, the adjustment power decreases synchronously to avoid over-adjustment. Through this dynamic adjustment method, the stability of the temperature in the optical measurement area is effectively guaranteed, providing environmental support for the accuracy of the scattered signal.

[0056] S3. Adaptive acquisition and interference suppression of scattered signals, as detailed below: S31. Under stable optical measurement conditions, a dual-detector array is used to collect the light signal scattered by the sample, and the detector gain is dynamically adjusted according to the real-time signal-to-noise ratio of the light signal, including: Real-time calculation of the signal-to-noise ratio of the signals acquired by the dual detector groups; The dual detector array is deployed according to the scattering angle. The central detector array focuses on collecting small-angle scattered light and is installed in the paraxial region of the laser incident direction, adapting to small-angle scattering signals dominated by small-diameter particles. The edge detector array is arranged in a ring around the central detector and is responsible for collecting large-angle scattered light, corresponding to large-diameter particles or samples with complex particle size distributions. The two detector arrays start collecting data synchronously, with the acquisition frequency consistent with the temperature and shading monitoring frequencies mentioned earlier. This ensures that the light signal corresponds one-to-one with other environmental and control parameters in the time dimension, laying the foundation for multi-dimensional data collaborative analysis.

[0057] During the acquisition process, the raw optical signal is preprocessed by using a signal filtering algorithm to eliminate circuit noise and ambient light interference, while preserving the characteristic contours of the effective scattered light signal. When calculating the signal-to-noise ratio (SNR) in real time, the average amplitude of the effective scattered signal is used as the signal intensity, and the standard deviation of the background signal amplitude acquired by the detector in the absence of a sample is used as the noise intensity. The SNR calculation formula is as follows: ; in The signal-to-noise ratio of the signal acquired by a single group of detectors. The average amplitude of the effective scattered signal, This represents the standard deviation of the background noise amplitude. The signal-to-noise ratio (SNR) of each of the two detector groups is calculated independently. Due to the natural difference in the intensity of scattered light at small and large angles, the SNR of the two groups may differ, requiring separate gain adjustments to avoid distortion of the detector signal caused by a single gain parameter.

[0058] When the signal-to-noise ratio (SNR) is below the preset lower limit, the control gain is adjusted increasing according to the logarithmic function of the ratio of SNR to the preset lower limit; when the SNR is above the preset upper limit, the control gain is adjusted decreasing according to the logarithmic function of the ratio of SNR to the preset upper limit.

[0059] The preset signal-to-noise ratio (SNR) upper and lower limits are determined based on the detector's performance parameters and measurement accuracy requirements. The preset lower limit ensures that weak signals can be effectively identified and the noise ratio is controllable, while the preset upper limit is set to prevent detector saturation caused by strong signals. The two sets of detectors can set different SNR thresholds according to their own signal strength characteristics. This embodiment uses a logarithmic function to adjust the gain. Its core advantage lies in the wide dynamic range of the SNR. Logarithmic adjustment can achieve smooth adaptation within a wide range, avoiding signal distortion caused by sudden gain changes at low SNR and preventing signal detail loss caused by sudden gain drops at high SNR.

[0060] When the signal-to-noise ratio (SNR) is below a preset lower limit, the gain is adjusted incrementally according to the logarithmic function of the ratio of SNR to the preset lower limit. The calculation formula is as follows: ; in This is the detector gain after incremental adjustment. This is the initial value of the gain. A lower limit is preset for the signal-to-noise ratio. This is the gain adjustment ratio coefficient, the value of which is determined based on the detector's gain adjustment range and signal sensitivity calibration. This formula controls the gain increment rate through a logarithmic relationship; the lower the signal-to-noise ratio, the smoother the gain increment, avoiding excessive gain amplification of noise.

[0061] When the signal-to-noise ratio (SNR) is higher than the preset upper limit, the gain is adjusted decreasing according to the logarithmic function of the ratio of SNR to the preset upper limit. The calculation formula is as follows: ; in To decrease the adjusted detector gain, Set an upper limit for the signal-to-noise ratio. To decrease the adjustment ratio coefficient, and It can be independently calibrated to adapt to the detector's attenuation requirements for strong signals. During adjustment, the gain value is always limited within the detector's rated gain range to prevent damage to the detector or abnormal signal acquisition caused by exceeding the range.

[0062] Each of the two detector groups executes independent gain adjustment logic. The central detector group, targeting the strong signal characteristics of small-angle scattered light, focuses on avoiding saturation by decreasing the gain; while the edge detector group, targeting the weak signal characteristics of large-angle scattered light, focuses on improving signal recognition by increasing the gain, ultimately achieving high-quality acquisition of scattered light signals from all angles.

[0063] S32. Perform detector self-calibration and multi-channel signal cross-verification before and after acquisition, respectively, and output effective scattering signals, including: Before each signal acquisition, the dual detector array is illuminated with a standard light source to generate calibration coefficients.

[0064] The self-calibration operation is initiated before each scattered light signal acquisition. Its core function is to provide a reference signal of known intensity using a standard light source to calibrate the response consistency of the dual detector array, eliminating detector sensitivity drift caused by long-term use or environmental changes. The standard light source is a monochromatic light source with stable intensity and the same wavelength as the measured laser. Its output light intensity has been metrologically calibrated and has a traceable reference value, ensuring calibration accuracy.

[0065] During calibration, a standard light source illuminates the central detector group and the edge detector group at preset angles, with each angle corresponding to a specific range of angles in the subsequent scattered light acquisition, simulating the actual incident state of scattered light. The detectors acquire the light signals from the standard light source, and combined with the known light intensity of the standard light source, calculate the calibration coefficient for each detector group. The formula for calculating the calibration coefficient is: ; in For a single group of detectors, the calibration coefficients are... The known reference light intensity of the standard light source. This represents the standard light intensity collected by the detector. The central detector group and the edge detector group each independently calculate calibration coefficients for correcting subsequent acquired signals. These calibration coefficients are updated with each self-calibration before each acquisition to ensure compatibility with the detector's real-time operating status.

[0066] After calibration, the standard light source is automatically turned off, and the system switches to diffused light acquisition mode to avoid interference from the standard light source with subsequent measurement signals. If the calibration coefficient calculated during the calibration process exceeds the preset reasonable range, a prompt will be triggered, the acquisition process will be paused, and the detector or standard light source will be checked for faults to ensure the effectiveness of the calibration.

[0067] After each signal acquisition, the signals acquired by the dual detector groups are time-synchronized and correlation-analyzed to remove abnormal signals with correlation coefficients lower than a preset threshold, and the retained valid signals are corrected using calibration coefficients.

[0068] After signal acquisition is completed, time synchronization is performed first. Although the central and edge detector groups start acquiring data synchronously, slight time differences may exist due to variations in circuit response speed. Alignment and calibration based on the timestamps of the two sets of signals are necessary to ensure complete synchronization of the scattered light signals at corresponding angles in time, providing an accurate data foundation for subsequent correlation analysis. Time synchronization is based on the system's unified clock, with corrections controlled within microseconds to avoid distortion of correlation judgments caused by time deviations.

[0069] After synchronization, correlation analysis is performed on the signals acquired by the dual detector groups. The core of this analysis involves calculating the correlation coefficient between the two sets of signals to determine their consistency and validity. The correlation analysis uses the Pearson correlation coefficient algorithm, and the calculation formula is as follows: ; in is the Pearson correlation coefficient of the dual detector group signal, with a value range of [-1, 1]. The signal value of the i-th sampling point of the central detector group. Let i be the signal value of the i-th sampling point of the edge detector group. The average value of the signal from the central detector group. The mean value of the signals from the edge detector group. This represents the total number of sampling points. The closer the correlation coefficient is to 1, the better the consistency between the two sets of signals and the less susceptible they are to abnormal interference.

[0070] The preset correlation coefficient threshold is determined based on measurement accuracy requirements and historical data calibration, and is typically set within a reasonable range close to 1 to ensure that only obviously abnormal signals are removed. When the calculated correlation coefficient is lower than the preset threshold, the signal is considered abnormal, possibly due to sudden environmental interference, momentary detector malfunction, etc., and the batch of synchronization signals is automatically removed. If the correlation coefficient is lower than the threshold multiple times consecutively, a fault diagnosis prompt is triggered to avoid continuous collection of invalid data.

[0071] The retained valid signals are corrected using the calibration coefficients obtained before acquisition to eliminate detector response bias. The correction formula is as follows: ; in The corrected effective scattering signal, The original signal retained after cross-validation. These are the calibration coefficients for the corresponding detector groups. The signals from the central detector group and the edge detector group are corrected using their respective calibration coefficients. After correction, they are integrated into a set of effective scattered signals at all angles for subsequent data inversion.

[0072] S4. Adaptive Inversion Calculation of Particle Size Distribution: Based on the effective scattering signal, real-time shading monitoring data, temperature monitoring data, and updated dispersion parameters, the particle size distribution of the sample under test is solved using an inversion algorithm, including: S41. Collect the updated dispersion parameters, adjusted laser intensity and sample flow rate, adjusted temperature control power, and dynamically adjusted detector gain as process parameters.

[0073] The collection of process parameters covers the entire process, including sample dispersion, optical measurement, and signal acquisition, ensuring that the inversion calculations fully reflect the actual state of the detection process. The collected parameters include dynamically adjusted ultrasonic power, stirring speed, dispersant dosage, adjusted laser intensity, sample flow rate, temperature control power, and independent gain values ​​for the central and peripheral detector groups.

[0074] All process parameters are aligned with the acquisition timestamps to ensure complete synchronization with the effective scattered signal in the time dimension, avoiding inversion errors caused by timing misalignment. To eliminate dimensional differences between different parameters, the process parameters need to be standardized using a linear normalization method, calculated as follows: ; in Let j be the standardized value of the j-th process parameter. The original value of the parameter. This is the historical minimum value of this parameter. This is the historical maximum value of the parameter. The standardized parameter will be fused with the effective scattered signal after weight allocation.

[0075] S42. Using the analytic hierarchy process (AHP), combined with real-time shading monitoring data and temperature monitoring data, adaptive weights are assigned to each process parameter.

[0076] Based on the analytic hierarchy process (AHP), the influence weights of each process parameter on the inversion results are dynamically adjusted according to real-time monitoring data, so that the weight allocation fits the current detection scenario. The hierarchical structure is divided into three layers: the target layer is to improve the accuracy of granular distribution inversion; the criterion layer contains real-time shading monitoring data and temperature monitoring data; and the scheme layer contains various process parameters collected in S41.

[0077] First, a judgment matrix is ​​constructed based on the criteria layer. The matrix elements are determined according to the deviation of real-time shading degree from temperature. When the shading degree deviates from the preset optimal range, the weight coefficients of process parameters related to concentration increase; when the temperature deviates from the set isothermal range, the weight coefficients of process parameters related to environmental stability increase. The formula for constructing the judgment matrix is:

[0078] in The criterion layer judgment matrix, The importance coefficient of shading degree relative to temperature. , Based on shading deviation Deviation from temperature calculate, , This is the adjustment coefficient.

[0079] Calculate the criterion layer weight vector using the eigenvector method. ,in As a weight of shading degree, The weights are determined by temperature. Then, a judgment matrix is ​​constructed relative to each criterion layer, and the weights of the scheme layer relative to the criterion layers are calculated using the same method. Finally, the adaptive weights of each process parameter are obtained by combining the criterion layer weights. .

[0080] To ensure the rationality of weight allocation, a consistency check is required. The formula for calculating the consistency index is as follows: ; in To determine the largest eigenvalue of a matrix, To determine the order of a matrix.

[0081] The formula for calculating the consistency ratio is: ; in This is the average random consistency index. When... If the judgment matrix satisfies the consistency requirement, the weight allocation is valid; otherwise, the judgment matrix is ​​reconstructed until the consistency condition is met.

[0082] S43. Input the weighted process parameters and the effective scattering signal into the inversion algorithm for iterative solution, and output the particle size distribution of the sample to be tested.

[0083] The weighted process parameters are fused with the effective scattering signal to construct the input vector for the inversion algorithm. The effective scattering signal is the preprocessed set of full-angle scattering signals. , The number of scattering angles; the weighted process parameters are: , The number of process parameters; the input vector is This enables deep fusion of signals and parameters.

[0084] The inversion algorithm is an improved bee colony artificial algorithm, and the improvements include: 1. The search step size is dynamically adjusted based on updated dispersion parameters and optical measurement parameters to ensure that the step size adapts to the current particle dispersion state and optical environment. Specifically, the step size calculation formula is as follows:

[0085] in For dynamic search step size, The initial step size, , For adjustment coefficients, The standardized mean of the dispersion parameter, This represents the standardized mean of the optical parameters. A better dispersion effect and a more stable optical environment allow for a smaller step size, enabling the algorithm to perform a finer search near the optimal solution; conversely, a larger step size improves the global search capability.

[0086] 2. By integrating real-time shading and temperature monitoring data, a fitness function with environmental parameter correction terms is constructed as follows: ; The fitness function is based on the deviation between the theoretical scattering signal and the actual effective scattering signal, and incorporates real-time shading and temperature monitoring data as environmental correction terms. For fitness value, This is the theoretical scattered signal value. , This is the environmental correction factor. For real-time shading, For optimal light blocking, For real-time temperature, To set a constant temperature. The smaller the fitness value, the closer the current particle size distribution solution is to the actual situation.

[0087] 3. Dynamically adjust the division of labor among bee colonies for single-peak, double-peak, and triple-peak distributions.

[0088] The bee colony includes mercenary bees, observation bees, and scout bees, with the division of labor dynamically adjusted according to the granularity distribution type. The distribution type is initially determined by the spectral characteristics of the effective scattered signal: a single-peaked distribution has a single peak, a double-peaked distribution has two peaks, and a triple-peaked distribution has three peaks.

[0089] In a unimodal distribution, hired bees account for 60%, observation bees for 30%, and scout bees for 10%, focusing on fine-grained local searches. In a bimodal distribution, hired bees account for 50%, observation bees for 40%, and scout bees for 10%, balancing local and global searches. When the distribution is three-peaked, the proportion of hired bees is 40%, the proportion of observation bees is 40%, and the proportion of scout bees is 20%, which enhances the global search capability and avoids missing peaks.

[0090] 4. Incorporate the weighted process parameters into the iterative optimization until the termination condition is met, and output the granularity distribution solution.

[0091] The weighted process parameters are incorporated into the new solution generation process of the hired bee. The new solution calculation formula is as follows:

[0092] in For the i-th hired bee, a new solution for the j-th dimension. For the solution of the k-th randomly selected hired bee in the j-th dimension, This is the standardized weighted value of the j-th process parameter.

[0093] During the iteration process, the observer bee selects food sources based on the fitness value of the hired bee, while the scout bee periodically eliminates long-unupdated food sources and randomly generates new solutions. The termination condition is reaching the preset maximum number of iterations. Or, the change in fitness value over several consecutive iterations is less than a preset error threshold. , Based on measurement accuracy requirements.

[0094] When the termination condition is met, the particle weight frequency distribution curve corresponding to the current optimal solution is output, and characteristic parameters such as median particle size, particle size range, and peak particle size are extracted to form the final particle size distribution result.

[0095] First, the bee colony size and iteration parameters are initialized, generating a batch of random granular distribution solutions as initial food sources. Simultaneously, the initial search step size, initial bee colony division of labor ratio, and termination condition parameters are determined. Upon entering the iteration phase, the hired bees generate new granular distribution solutions based on the dynamically adjusted search step size and weighted process parameters. The fitness values ​​of the new and old solutions are calculated using a fitness function with environmental parameter correction terms, and a greedy selection strategy is employed to retain the better solutions. Observer bees select food sources probabilistically based on their fitness values, repeating the process of new solution generation and greedy selection by the hired bees. Scout bees continuously monitor the improvement status of each food source, eliminating long-unupdated food sources and randomly generating new solutions to avoid the algorithm getting trapped in local optima. In each iteration, the bee colony division of labor ratio is dynamically adjusted based on the spectral characteristics of the scattering signal corresponding to the current optimal solution, adapting to possible single-peak, double-peak, or triple-peak distribution scenarios. The iteration process continues until a preset maximum number of iterations is reached or the change in fitness value across multiple consecutive iterations is less than an error threshold. At this point, the granular distribution solution corresponding to the current optimal fitness value is output, completing the entire optimization process.

[0096] On the other hand, embodiments of the present invention also disclose an intelligent particle size detection system with multi-parameter control, such as... Figure 3 As shown, it includes: The material characteristics and dispersion module is used to collect the material characteristics of the particles to be tested and match the initial dispersion parameters, mix them to form an initial suspension, dynamically update the parameters based on image recognition results and real-time shading data, and output a stable dispersion suspension. The optical measurement and control module is used to receive a stable dispersed suspension and introduce it into the optical measurement area. It adjusts the laser light intensity and sample flow rate based on the real-time shading degree and adjusts the temperature control power in combination with temperature data to form stable optical measurement conditions. The signal acquisition and processing module includes a dual detector group for acquiring scattered light signals, dynamically adjusting the gain according to the signal-to-noise ratio, performing self-calibration and cross-validation before and after acquisition, and outputting effective scattered signals. The data inversion module is used to collect effective scattering signals, monitoring data, and updated dispersion parameters, and solves the granularity distribution through an inversion algorithm.

[0097] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0098] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A multi-parameter controlled intelligent particle size detection method, characterized in that, Includes the following steps: The material characteristics of the particle sample to be tested are collected to match the initial dispersion parameters; the particle sample to be tested is mixed with the dispersion medium to form an initial suspension; based on the image recognition results of the initial suspension and the real-time shading monitoring data, the initial dispersion parameters are dynamically updated to obtain the suspension to be tested in a preset stable dispersion state. A suspension sample in a stable dispersed state is introduced into the optical measurement area. Based on real-time shading monitoring data, the intensity of the laser irradiating the suspension sample and the flow rate of the sample are adjusted. At the same time, the temperature control power is adjusted based on the temperature monitoring data of the optical measurement area to form stable optical measurement conditions. Under stable optical measurement conditions, the light signal scattered by the sample is collected using a dual detector group. The detector gain is dynamically adjusted according to the real-time signal-to-noise ratio of the light signal. Detector self-calibration and multi-channel signal cross-verification are performed before and after the acquisition, and an effective scattering signal is output. Based on the effective scattering signal, real-time shading monitoring data, temperature monitoring data, and updated dispersion parameters, the particle size distribution of the sample under test is solved by an inversion algorithm.

2. The intelligent particle size detection method with multi-parameter control according to claim 1, characterized in that, The material characteristics of the particle sample to be tested are collected to match the initial dispersion parameters, including: The reflectance spectrum characteristics, density, and surface charge properties of the particle sample to be tested are collected to construct a three-dimensional material feature vector; The three-dimensional material feature vector is matched with a pre-stored mapping database to output the corresponding initial dispersion parameter combination, which includes stirring speed, ultrasonic power, dispersant type and dosage.

3. The intelligent particle size detection method with multi-parameter control according to claim 1, characterized in that, Based on image recognition results and real-time shading monitoring data of the initial suspension, the initial dispersion parameters are dynamically updated to obtain the test suspension in a preset stable dispersion state, including: Extract the equivalent diameter of aggregates, the proportion of aggregates, and the particle uniformity from the image of the initial suspension; The real-time shading degree and shading degree change rate of the initial suspension were collected in conjunction with the data. Using the proportion of agglomerates and the degree of shading as feedback variables, a PID-PWM hybrid algorithm is adopted to gradient adjust the ultrasonic power, stirring speed and dispersant dosage according to preset priorities until the suspension state simultaneously meets the preset conditions of agglomerate proportion, particle uniformity, shading degree and shading degree change rate.

4. The intelligent particle size detection method with multi-parameter control according to claim 1, characterized in that, Based on real-time shading monitoring data, the laser intensity irradiating the suspension sample and the sample flow rate are adjusted, including: When the real-time shading degree is higher than the preset upper limit and the rate of change of shading degree is non-negative, the laser intensity is controlled to decrease exponentially; when the real-time shading degree is lower than the preset lower limit and the rate of change of shading degree is non-positive, the laser intensity is controlled to increase exponentially. Meanwhile, the sample flow rate is controlled to be linearly adjusted in proportion to the real-time shading value.

5. The intelligent particle size detection method with multi-parameter control according to claim 1, characterized in that, Adjusting the temperature control power based on monitoring data of the temperature in the optical measurement area to create stable optical measurement conditions includes: Real-time acquisition of temperature and temperature change rate in the optical measurement area; When the temperature deviates from the set constant temperature range by less than or equal to the first threshold, the temperature control module operates in low-power pulse mode; when the temperature deviates from the first threshold and the temperature change rate is greater than or equal to the second threshold, the temperature control power is adjusted proportionally to the product of the temperature deviation and the temperature change rate.

6. The intelligent particle size detection method with multi-parameter control according to claim 1, characterized in that, The optical signal scattered by the sample is acquired using a dual-detector array, and the detector gain is dynamically adjusted based on the real-time signal-to-noise ratio of the optical signal, including: Real-time calculation of the signal-to-noise ratio of the signals acquired by the dual detector groups; When the signal-to-noise ratio (SNR) is below the preset lower limit, the control gain is adjusted increasing according to the logarithmic function of the ratio of SNR to the preset lower limit; when the SNR is above the preset upper limit, the control gain is adjusted decreasing according to the logarithmic function of the ratio of SNR to the preset upper limit. The dual detector array includes a central detector array for collecting small-angle scattered light and an edge detector array for collecting large-angle scattered light.

7. The intelligent particle size detection method with multi-parameter control according to claim 1, characterized in that, Detector self-calibration and multi-channel signal cross-verification are performed before and after data acquisition to output effective scattered signals, including: Before each signal acquisition, the dual detector group is illuminated with a standard light source to generate calibration coefficients; After each signal acquisition, the signals acquired by the dual detector groups are time-synchronized and correlation-analyzed to remove abnormal signals with correlation coefficients lower than a preset threshold, and the retained valid signals are corrected using calibration coefficients.

8. The intelligent particle size detection method with multi-parameter control according to claim 1, characterized in that, Based on the effective scattering signal, real-time shading monitoring data, temperature monitoring data, and updated dispersion parameters, the particle size distribution of the sample under test is solved using an inversion algorithm, including: The updated dispersion parameters, adjusted laser intensity and sample flow rate, adjusted temperature control power, and dynamically adjusted detector gain were collected as process parameters. The analytic hierarchy process was used to assign adaptive weights to each process parameter by combining real-time shading monitoring data and temperature monitoring data. The weighted process parameters and the effective scattering signal are input together into the inversion algorithm for iterative solution, and the particle size distribution of the sample to be tested is output.

9. The intelligent particle size detection method with multi-parameter control according to claim 8, characterized in that, The inversion algorithm is an improved bee colony artificial algorithm, including: The search step size is dynamically adjusted based on updated dispersion parameters and optical measurement parameters; By integrating real-time shading and temperature monitoring data, a fitness function with environmental parameter correction terms is constructed. The division of labor among bee colonies is dynamically adjusted based on unimodal, bimodal, and trimodal distributions. The weighted process parameters are incorporated into the iterative optimization until the termination condition is met, and the granularity distribution solution is output.

10. A multi-parameter-controlled intelligent particle size detection system, used to implement the multi-parameter-controlled intelligent particle size detection method as described in any one of claims 1-9, characterized in that, include: The material characteristics and dispersion module is used to collect the material characteristics of the particles to be tested and match the initial dispersion parameters, mix them to form an initial suspension, dynamically update the parameters based on image recognition results and real-time shading data, and output a stable dispersion suspension. The optical measurement and control module is used to receive a stable dispersed suspension and introduce it into the optical measurement area. It adjusts the laser light intensity and sample flow rate based on the real-time shading degree and adjusts the temperature control power in combination with temperature data to form stable optical measurement conditions. The signal acquisition and processing module includes a dual detector group for acquiring scattered light signals, dynamically adjusting the gain according to the signal-to-noise ratio, performing self-calibration and cross-validation before and after acquisition, and outputting effective scattered signals. The data inversion module is used to collect effective scattering signals, monitoring data, and updated dispersion parameters, and solves the granularity distribution through an inversion algorithm.

Images