A method and system for quality control in flour production

Abstract

The application relates to the flour production technical field and discloses a quality control method and system for flour production, which comprises the following modules: a data acquisition module, a disturbance level determination module, a quality fluctuation determination module and a monitoring mode switching module.The data acquisition module comprises a plurality of aerosol particle detectors arranged in a workshop and is used for collecting the particle size distribution and concentration change of dust particles in the air during flour production.The disturbance level determination module is used for calculating the disturbance level under the current process parameters based on the fluctuation degree of the particle size distribution and the concentration change.The quality fluctuation determination module is used for calculating the quality fluctuation value under the current batch flour production state according to the disturbance level and the particle size distribution.The monitoring mode switching module is configured to perform mode switching judgment according to the comparison result of the quality fluctuation value and a preset threshold value.The application realizes disturbance source identification and dynamic control strategy linkage through the particle size distribution and concentration change of aerosol particles, and improves the consistency of the finished product quality.

Description

Technical Field

[0001] This invention relates to the field of flour production technology, and more specifically, to a quality control method and system for flour production. Background Technology

[0002] As a crucial basic grain raw material, flour's quality stability directly affects the taste, texture, and nutritional properties of downstream food products. Flour quality depends not only on the wheat raw material itself but also on a combination of process factors, including the uniformity of grinding, screening efficiency, and particle size distribution during milling.

[0003] In existing flour production processes, although some factories have introduced automated equipment to link the crushing and screening processes, these control systems largely rely on the operating parameters of the crushing and screening equipment as indirect indicators, lacking the ability to perceive the real-time distribution of powder during actual production. Flour, as a fine powder system, typically exhibits particle size fluctuations during production as slight structural shifts, asymmetrical distribution, or tail-end accumulation; these characteristics are difficult to accurately identify using traditional power, current, or speed feedback. Furthermore, the particle size and concentration characteristics of airborne dust particles are highly correlated with the uniformity of crushing and screening efficiency during production; however, traditional quality control has not established a dynamic correlation mechanism between dust aerosol characteristics and flour quality. Especially under conditions of micro-disturbance, such as screen clogging, fluctuations in crushing load, or changes in raw material moisture, the system often lacks early detection capabilities, making it difficult to correct particle size distribution shifts in a timely manner, ultimately leading to increased fluctuations in finished product quality between batches.

[0004] Furthermore, the current flour production quality monitoring system still suffers from problems such as lagging feedback mechanisms and crude control strategies, lacking the ability to diagnose disturbances and adaptively control processes based on real-time data. For example, before particle size distribution significantly affects the quality indicators of the finished product, the system often fails to identify potential risks in advance and has failed to establish effective process parameter adjustment and intervention paths. Summary of the Invention

[0005] In view of this, the present invention proposes a quality control method and system for flour production to solve the problems of existing technologies that cannot accurately reflect the impact of particle size distribution structure shift on finished product quality and slow response of control path.

[0006] On one hand, the present invention proposes a quality control system for flour production, comprising:

[0007] The data acquisition module includes several aerosol particle detectors deployed in the workshop. The data acquisition module is configured to collect the particle size distribution and concentration changes of dust particles in the air during flour production.

[0008] The disturbance level determination module, connected to the data acquisition module, is configured to calculate the disturbance level under the current process parameters based on the fluctuation degree of particle size distribution and concentration change.

[0009] The quality fluctuation determination module, connected to the disturbance level determination module, is configured to calculate the quality fluctuation value of the current batch of flour under the production state based on the disturbance level and particle size distribution.

[0010] The monitoring mode switching module is configured to make a mode switching judgment based on the comparison result of the quality fluctuation value and the preset threshold. The mode switching judgment includes: when the quality fluctuation value is lower than the first threshold, maintaining the current working mode; when it is equal to or higher than the first threshold and lower than the second threshold, entering the enhanced monitoring mode, which includes adjusting the acquisition frequency according to the relationship between the quality fluctuation value and the second threshold; and when it is equal to or higher than the second threshold, entering the control mode to eliminate flour particle size fluctuation.

[0011] Furthermore, the disturbance level is obtained by weighted summation based on the standard deviation of particle size distribution and concentration change; during weighted summation, the contribution value calculated by the SHAP model is set as the weight coefficient.

[0012] Furthermore, the step of obtaining the quality fluctuation value includes:

[0013] Obtain the disturbance level sequence within the acquisition period; assign the concentration change value to 0, construct the disturbance level standard sequence based on the standard particle size distribution of dust particles and the concentration change value assigned to zero, and obtain the maximum difference between the disturbance level sequence and the disturbance level standard sequence at the same time within the acquisition period, which is recorded as the disturbance fluctuation amplitude.

[0014] A particle size distribution curve is established based on the particle size distribution within the collection period. The particle size distribution curve is then divided into several particle size intervals. The difference between the average value of the particle size distribution in each particle size interval and the reference value is calculated. The summation yields the structural offset.

[0015] The mass fluctuation value is calculated using a logarithmic function based on the disturbance fluctuation amplitude and the structural offset.

[0016] Furthermore, the sampling frequency in the enhanced monitoring mode is determined by the following method:

[0017] The maximum sampling frequency of the aerosol particle detector is obtained, wherein the sampling frequency is (mass fluctuation value / second threshold) × maximum sampling frequency.

[0018] Furthermore, the monitoring mode switching module also includes a disturbance factor tracing unit, which is configured as follows:

[0019] By comparing the trend of the main peak position of the particle size distribution curve with the proportion of the width expansion on both sides of the particle size distribution curve at the main peak position within several consecutive collection cycles, it is determined whether there is a phenomenon of particle size centroid shift and particle size structure distribution asymmetry, and adjustments are made based on the judgment results.

[0020] When both phenomena occur simultaneously, reduce the drive frequency of the crushing motor and the screening speed.

[0021] When only particle size centroid shift exists, reduce the drive frequency of the crushing motor;

[0022] When only the particle size distribution is asymmetrical, reduce the screening speed.

[0023] Furthermore, the monitoring mode switching module also includes a disturbance trend monitoring unit, which is configured as follows:

[0024] After the disturbance factor tracing unit ends the regulation, the particle size distribution curve of the corresponding acquisition period when the regulation is performed is analyzed. If the main peak position of the particle size distribution curve shifts back and the width expansion ratio of the particle size distribution curve returns, it is recorded as effective regulation; otherwise, it is judged as ineffective regulation.

[0025] Furthermore, when the control is deemed ineffective, the disturbance level and structural offset are obtained. If the disturbance level does not decrease within two consecutive acquisition cycles and the structural offset increases, the crushing motor frequency is kept constant and the screening speed is further reduced until the particle size distribution curve width expansion ratio stops increasing.

[0026] Furthermore, when a regulation is deemed ineffective, it also includes:

[0027] Analyze the changes in particle size distribution curves over two consecutive sampling cycles after adjusting the screening speed. When an increase in structural offset is detected, determine whether the disturbance source is a fluctuation in powder load based on the following conditions:

[0028] If, within two consecutive sampling cycles after the screening speed adjustment, the dust concentration change exceeds 10% of the previous sampling and the number of times it exceeds the preset number, it is determined that the concentration fluctuates frequently.

[0029] If, within two consecutive sampling periods after the adjustment of the screening speed, the absolute value of the maximum change in particle size variation rate is greater than the baseline variation rate, it is determined that the particle size variation rate exhibits oscillation characteristics.

[0030] When the concentration fluctuates frequently and exhibits oscillation characteristics, the current disturbance is determined to be a fluctuation in powder load, and the drive frequency of the crushing motor and the screening speed are adjusted down simultaneously.

[0031] Furthermore, after simultaneously reducing the drive frequency of the crushing motor and the screening speed, the particle size distribution curve characteristics of subsequent sampling cycles were monitored.

[0032] If, in any period, the main peak of the particle size distribution curve moves toward the main peak of the standard particle size distribution curve, and the number of times the dust concentration change exceeds 10% of the previous sampling is less than or equal to three, it is considered an effective control.

[0033] If the main peak does not shift back, determine whether there is a particle size accumulation region. When the structural displacement in a certain particle size range is greater than the preset value of the structural displacement in the adjacent particle size range, and there is no main peak in the current particle size range, then there is a particle size accumulation region. Maintain the current crushing motor drive frequency and screening speed unchanged until the particle size accumulation tends to disappear.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] By deploying aerosol particle detectors, it is possible to continuously collect the distribution and concentration changes of dust particles in the air during flour production, enabling online sensing of particle size fluctuation trends and overcoming the limitations of traditional monitoring methods that rely on indirect parameters such as motor power and current.

[0036] By constructing perturbation levels and quality fluctuation values, and combining them with the standard deviation, interval offset, and trend of particle size structure, we can accurately identify complex perturbation features such as particle size centroid shift, asymmetric distribution, and structural packing, thereby improving the detection sensitivity of particle size structure anomalies.

[0037] The system is equipped with three monitoring modes and responds to different disturbance intensities by classifying quality fluctuation values. This enables adaptive adjustment of the acquisition frequency and timely switching of control strategies, thereby improving the dynamism and foresight of quality monitoring.

[0038] Through the disturbance factor tracing and trend monitoring unit, the attributes of the disturbance source can be accurately determined based on the main peak shift, width change and particle size range accumulation phenomenon of the particle size curve. This guides the linkage or graded adjustment of the crushing motor drive frequency and the screening speed, avoiding over-adjustment or response lag.

[0039] This system establishes a real-time correlation between dust aerosol characteristics and powder particle size, forming a complete closed-loop process of diagnosis, identification, and control. This effectively suppresses quality deviations caused by batch-to-batch particle size fluctuations and process interference, significantly improving the stability of the flour production process and the consistency of the finished product.

[0040] On the other hand, the present invention proposes a quality control method for flour production, comprising:

[0041] S1: Collect the particle size distribution and concentration changes of airborne dust particles during flour production by deploying several aerosol particle detectors in the workshop;

[0042] S2: Calculate the disturbance level under the current process parameters based on the fluctuation of particle size distribution and concentration changes;

[0043] S3: Calculate the quality fluctuation value of the current batch of flour under the production status based on the disturbance level and particle size distribution;

[0044] S4: Based on the comparison result between the quality fluctuation value and the preset threshold, perform mode switching judgment; and eliminate flour particle size fluctuation.

[0045] It should be noted that the quality control method and system for flour production of the present invention have the same beneficial effects, and will not be described in detail here. Attached Figure Description

[0046] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:

[0047] Figure 1 This is a functional block diagram of a quality control system for flour production provided in an embodiment of the present invention.

[0048] Figure 2 A flowchart of a quality control method for flour production provided in an embodiment of the present invention. Detailed Implementation

[0049] Exemplary embodiments of the present application will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0050] See Figure 1 As shown, this embodiment of the invention provides a quality control system for flour production, comprising:

[0051] The data acquisition module includes several aerosol particle detectors deployed in the workshop. The data acquisition module is used to collect the particle size distribution and concentration changes of dust particles in the air during flour production.

[0052] The disturbance level determination module is connected to the data acquisition module. Based on the fluctuation degree of particle size distribution and concentration change, it calculates the disturbance level under the current process parameters.

[0053] The quality fluctuation determination module is connected to the disturbance level determination module and calculates the quality fluctuation value of the current batch of flour under the production status based on the disturbance level and particle size distribution.

[0054] The monitoring mode switching module is configured to make mode switching judgments based on the comparison results of the quality fluctuation value and the preset threshold. The mode switching judgments include: when the quality fluctuation value is lower than the first threshold, maintaining the current working mode; when it is equal to or higher than the first threshold but lower than the second threshold, entering the enhanced monitoring mode, which includes adjusting the acquisition frequency according to the relationship between the quality fluctuation value and the second threshold; and when it is equal to or higher than the second threshold, entering the control mode to eliminate flour particle size fluctuations.

[0055] It should be noted that this embodiment uses aerosol particle size and concentration characteristics as external response signals of powder state to construct a highly sensitive sensing mechanism related to changes in flour particle size distribution. Based on this, a multi-level feedback loop is established, encompassing disturbance identification, quality fluctuation judgment, mode switching, and process control. This system utilizes the behavior of dust particles inevitably generated during production to infer the stability of the crushing and screening processes, thereby achieving real-time monitoring and dynamic control of quality status without interrupting production.

[0056] In this embodiment, the aerosol particle detector includes a laser particle size analyzer, a PM sensor, a constant sampling flow controller, and an environmental compensation sensor. During flour production, powder particles are continuously released during the crushing and screening processes. The particle size distribution and concentration levels of these particles in space change with fluctuations in the process conditions. Therefore, collecting this data can be considered a non-contact projection monitoring of the production status.

[0057] The disturbance level is based on the standard deviation of particle size distribution and the absolute difference in concentration variation. The former reflects the stability of screening and classification effects, while the latter reflects the volatility of powder loading. The two are weighted and superimposed to form a unified index representing the stability of the current process state. This index can be used to determine whether the current system is experiencing minor disturbances, periodic oscillations, or structural shifts.

[0058] Even minor shifts in flour particle size distribution, such as main peak displacement or tail accumulation, can cause changes in the finished flour's water absorption, gluten strength, and baking performance. This module evaluates the current disturbance level in conjunction with the degree of particle size distribution shift, outputting a quality fluctuation value that characterizes the production status. Essentially, this value measures the degree to which the current flour particle size distribution deviates from the standard structure and can serve as a trigger for subsequent control actions.

[0059] In some embodiments of this application, the perturbation level is obtained by weighted summation based on the standard deviation of particle size distribution and concentration change; during weighted summation, the contribution value calculated by the SHAP model is used as the weighting coefficient.

[0060] It should be noted that, firstly, real-time data from several aerosol particle detectors within the workshop is collected, including: a sequence of dust particle size distribution data within a unit sampling period, and the corresponding dust concentration change values ​​at different time points. The particle size distribution data is used to calculate the standard deviation of the particle size distribution, and the concentration change value reflects the dynamic fluctuation of the powder release rate. Secondly, the two core indicators required for the disturbance level are quantified: the standard deviation of the particle size distribution is calculated statistically by statistically analyzing the particle size frequency distribution within the current sampling period; the dust concentration change value is represented by the change in the ratio of the average concentration between adjacent sampling periods, reflecting the instantaneous impact of the current equipment operation on powder release.

[0061] Then, to obtain the weight relationship between the two indicators in the perturbation level, a dataset was constructed based on historical sampling data, and a prediction function for the perturbation level was obtained through regression model training. In this process, the SHAP model was introduced to perform feature contribution analysis on the model results, calculating the average contribution values ​​of the particle size distribution standard deviation and concentration change in the historical samples. These contribution values ​​were used as weighting coefficients and assigned to the two indicators respectively.

[0062] Finally, a weighted summation method is used to synthesize the current disturbance level value. This involves multiplying the standard deviation of particle size distribution and the concentration change value by their respective weighting coefficients and then summing the results to obtain the disturbance level value, which is used for subsequent fluctuation judgment and control logic switching.

[0063] In some embodiments of this application, the step of obtaining the quality fluctuation value includes:

[0064] Obtain the disturbance level sequence within the acquisition period; assign the concentration change value to 0, construct the disturbance level standard sequence based on the standard particle size distribution of dust particles and the concentration change value assigned to zero, and obtain the maximum difference between the disturbance level sequence and the disturbance level standard sequence at the same time within the acquisition period, which is recorded as the disturbance fluctuation amplitude.

[0065] A particle size distribution curve is established based on the particle size distribution within the collection period. The particle size distribution curve is then divided into several particle size intervals. The difference between the average particle size distribution value and the reference value in each particle size interval is calculated, and the summation yields the structural offset.

[0066] The mass fluctuation value is calculated using a logarithmic function based on the disturbance fluctuation amplitude and the structural offset.

[0067] Specifically, the steps for obtaining the quality fluctuation value include:

[0068] Step 1: Collect the disturbance level data sequence output by the disturbance level determination module within the current sampling period, denoted as D={d1,d2,…,dn}, where each disturbance level value di (i is a positive integer) is calculated by weighting the standard deviation of particle size distribution and concentration change, reflecting the overall disturbance degree of the powder state at the current moment.

[0069] Step 2: To obtain the perturbation deviation relative to the ideal state, define the standard perturbation level sequence D. * ={d1 * ,d2 * ,…,dn *}, di * This represents the standard perturbation level, where i is a positive integer. In this process, all concentration variation terms are set to 0, the standard particle size distribution parameters are retained, and the perturbation level is recalculated as a standard reference. This sequence reflects the particle size perturbation intensity under ideal conditions with no concentration fluctuations.

[0070] Step 3: Calculation of disturbance fluctuation amplitude

[0071] Calculate sequences D and D * The absolute value of the difference at each time point within the same sampling period is selected, and the largest difference is taken as the disturbance fluctuation amplitude, denoted as:

[0072]

[0073] This value reflects the strongest deviation of the current powder disturbance from the ideal state.

[0074] Step 4: Extract all particle size distribution data within the sampling period, construct a particle size distribution curve, and divide the particle size range into several equal-width particle size intervals {I1, I2, ..., Im}. Within each particle size interval, calculate the mean actual distribution frequency fi and the corresponding reference standard value fi. * The difference between them, when summed, yields the structural offset S:

[0075]

[0076] Among them, the reference value fi * The value is derived from the standard particle size distribution curve, where i represents a positive integer.

[0077] Step 5: Based on the disturbance fluctuation amplitude A and the structural offset S, the following coupling method is used:

[0078] Q = ln(1 + A·S);

[0079] Here, Q represents the final quality fluctuation value, which reflects the combined effects of disturbance intensity and structural changes, and can more comprehensively characterize the quality stability of the powder produced in the current batch.

[0080] In some embodiments of this application, the sampling frequency in enhanced monitoring mode is determined by the following method:

[0081] The maximum acquisition frequency of the aerosol particle detector is obtained. The acquisition frequency is the product of the ratio of the mass fluctuation value to the second threshold and the maximum acquisition frequency.

[0082] In this embodiment, firstly, the system obtains the maximum sampling frequency supported by the deployed aerosol particle detector. This frequency is the upper limit that the detector can achieve while maintaining detection accuracy and response stability, and is denoted as the maximum sampling frequency. Secondly, the system calls the quality fluctuation judgment module to obtain the quality fluctuation value calculated within the current sampling period. This quality fluctuation value characterizes the degree of particle size disturbance and the magnitude of distribution structure change under the current flour production state. Then, the system presets a reference threshold for mode switching judgment, serving as the boundary between enhanced monitoring and control modes, denoted as the second threshold. This threshold represents the system's upper limit tolerance for acceptable particle size fluctuation amplitude. Based on this, the system calculates the ratio between the current quality fluctuation value and the second threshold to obtain the relative intensity of the current quality disturbance within the system's acceptable range. According to this ratio, the maximum sampling frequency is scaled proportionally to obtain the sampling frequency corresponding to the current sampling period. Specifically, the current sampling frequency is equal to the ratio of the quality fluctuation value to the second threshold multiplied by the maximum sampling frequency. Finally, to prevent abnormal fluctuations in the sampling frequency from affecting system stability, boundary values ​​between the minimum and maximum sampling frequencies can be set during the calculation process to ensure that the final sampling frequency always fluctuates within a reasonable range.

[0083] In some embodiments of this application, the monitoring mode switching module further includes a disturbance factor tracing unit, which is configured to:

[0084] By comparing the trend of the main peak position of the particle size distribution curve with the proportion of the width expansion on both sides of the particle size distribution curve at the main peak position within several consecutive collection cycles, it is determined whether there is a phenomenon of particle size centroid shift and particle size structure distribution asymmetry, and adjustments are made based on the judgment results.

[0085] When both phenomena occur simultaneously, reduce the drive frequency of the crushing motor and the screening speed.

[0086] When only particle size centroid shift exists, reduce the drive frequency of the crushing motor;

[0087] When only the particle size distribution is asymmetrical, reduce the screening speed.

[0088] It should be noted that the monitoring mode switching module is equipped with a disturbance factor tracing unit. This unit is used to determine the dominant disturbance factor of the current flour particle size fluctuation based on the dynamic evolution characteristics of particle size distribution after the system enters the control mode, and trigger a differentiated control strategy accordingly.

[0089] Specifically, the disturbance factor tracing unit takes the dust particle size distribution curves obtained in multiple consecutive sampling periods as the analysis object, tracks the main peak position of the particle size distribution in each sampling period, obtains the changing trend of the main peak position on the time axis, and calculates the expansion ratio of the particle size distribution width on both sides of the main peak position, thereby constructing a disturbance structure judgment index.

[0090] The determination of the trend of the main peak position includes the following steps: First, determine the particle size range corresponding to the maximum value point of the particle size distribution curve in each sampling period, which is taken as the main peak position of that period; then analyze the displacement direction and displacement amplitude of the main peak position in the continuous sampling period. If the main peak shows a continuous deviation from the reference particle size (such as the standard particle size set by the process) in two or more periods, and the direction of deviation remains consistent, it can be determined that there is a particle size centroid shift phenomenon.

[0091] On the other hand, to assess whether the particle size distribution is symmetrical, the system divides symmetrical intervals into smaller and larger particle sizes based on the position of the main peak, and calculates the area, width, or particle size concentration of the intervals on both sides of the main peak. If the expansion on either side is significantly higher than that on the other side, and exceeds the set asymmetry threshold, then the particle size distribution can be judged to be asymmetrical.

[0092] The disturbance factor tracing unit performs classification and control through parallel analysis of the above two dimensions, combined with the following logical judgment: When particle size centroid shift and structural distribution asymmetry are detected simultaneously, that is, the main peak position continues to shift and the expansion ratio on both sides of the main peak is significantly different, it indicates that the current disturbance may originate from the particle size generation shift in the crushing process and the structural disturbance in the screening process. At this time, the system jointly lowers the driving frequency of the crushing motor and the rotation speed of the screening unit to achieve coordinated control of the particle size input and output ends.

[0093] When only the particle size centroid shift is detected while the structural symmetry remains stable, it indicates that the main disturbance comes from the crushing process. In this case, the system only reduces the drive frequency of the crushing motor to reduce the degree of raw material crushing per unit time, thereby suppressing the particle size shift trend.

[0094] When there is only structural asymmetry and the position of the main peak remains stable, it indicates that the disturbance may be caused by problems such as sieve hole blockage, decreased sieve efficiency, or unbalanced distribution during the screening process. In this case, the system only adjusts the screening speed to slow down the screening process and improve the particle size screening structure.

[0095] In addition, to avoid misjudgment caused by instantaneous fluctuations, the disturbance factor tracing unit is equipped with a periodic consistency judgment mechanism. Only when the main peak displacement trend or structural asymmetry shows consistency in at least two consecutive sampling periods will it be judged as a stable disturbance and trigger the corresponding control operation.

[0096] By configuring this module, the causes of particle size fluctuations can be analyzed during the process, which helps to intervene at the source, thereby improving the response speed and accuracy of particle size control and enhancing the system's ability to identify and adjust flour quality fluctuations early.

[0097] In some embodiments of this application, the monitoring mode switching module further includes a disturbance trend monitoring unit, which is configured to:

[0098] After the disturbance factor tracing unit finishes its regulation, the particle size distribution curve of the corresponding acquisition period during the regulation is analyzed. If the main peak position of the particle size distribution curve shifts back and the width expansion ratio of the particle size distribution curve returns, it is recorded as effective regulation; otherwise, it is judged as ineffective regulation.

[0099] It should be noted that the monitoring mode switching module includes a disturbance trend monitoring unit. This unit is used to verify and attribute the control effect after completing a disturbance control operation, thereby supporting whether to continue, maintain, or change the control strategy. Its implementation steps are as follows:

[0100] Step 1: Mark and adjust the initial sampling period

[0101] When the disturbance factor tracing unit executes the control command (such as adjusting the frequency of the crushing motor or the speed of the screening), the system records the sampling cycle number at this time as the control start point, which is denoted as the "controlled cycle".

[0102] Step 2: Obtain the particle size distribution curves for the regulated period and several subsequent periods.

[0103] Starting from the "controlled period", particle size distribution data for several consecutive sampling periods are acquired, and the corresponding particle size distribution curve is plotted based on the data of each period.

[0104] Step 3: Calculate the change in the position of the main peak

[0105] For each particle size distribution curve, the system extracts its main peak position (i.e., the particle size range corresponding to the maximum particle size distribution value) and compares it with the main peak position before adjustment.

[0106] If the position of the main peak moves from the deviation state before adjustment towards the reference standard particle size and maintains this trend, it is recorded as "main peak position shifted back"; otherwise, it is recorded as "main peak position did not shift back".

[0107] Step 4: Evaluate the width expansion ratio of the particle size distribution curve

[0108] Calculate the expansion ratio of particle size distribution width before and after regulation. The expansion ratio is calculated as follows: the sum of the widths of the distribution intervals on both sides of the main peak particle size is divided by the main peak intensity value, which reflects whether the curve is flat and whether the tail expands outward.

[0109] If the expansion ratio of the curve after adjustment decreases compared to before adjustment (i.e., it becomes more concentrated and the distribution tends to be symmetrical), it is considered as "width expansion ratio pullback"; otherwise, it is judged as "no pullback".

[0110] Step 5: Jointly assess the effectiveness of regulation

[0111] If the following two conditions are met simultaneously within two consecutive sampling periods after the adjustment: the main peak position of the particle size distribution curve shifts back; the width expansion ratio of the particle size distribution curve rebounds; then the system records this adjustment as "effective adjustment" and continues to observe or enters a stable monitoring state; otherwise, if either of the above conditions is not met, the adjustment is determined to be "ineffective adjustment" and submitted to other modules for further processing.

[0112] In some embodiments of this application, when it is determined that the control is invalid, the disturbance level and the structural offset are obtained. When the disturbance level does not decrease within two consecutive acquisition cycles and the structural offset increases, the frequency of the crushing motor is kept unchanged and the screening speed is further reduced until the width expansion ratio of the particle size distribution curve stops increasing.

[0113] It should be noted that, in this embodiment, the processing steps when an invalid control is determined to be invalid are as follows:

[0114] Step 1: Initiate the ineffective control identification process

[0115] After the disturbance factor tracing unit completes the adjustment operation (such as adjusting the crushing motor frequency and / or screening speed), the monitoring mode switching module analyzes the adjusted particle size distribution curve through the disturbance trend monitoring unit.

[0116] If the shift of the main peak position of particle size is not detected, or the proportion of distribution width expansion does not show a downward trend, then this round of regulation is determined to be "ineffective regulation" and proceeds to the next stage of processing.

[0117] Step 2: Extract the disturbance level and structural offset for two consecutive periods.

[0118] After determining "ineffective control", the system sequentially obtains the following two parameters for the current cycle and the previous cycle: Disturbance level: obtained by weighted summation of the standard deviation of particle size distribution and dust concentration change; Structural offset: reflects the degree of offset between the average value of each interval of particle size distribution and the reference standard.

[0119] Step 3: Determine if the disturbance level has not decreased continuously.

[0120] Compare the disturbance levels in the two current cycles: if the disturbance levels in the two cycles remain unchanged or increase, it is determined that the disturbance has not weakened; if they decrease, the control action is temporarily suspended and observation continues.

[0121] Step 4: Determine if the structural offset has increased.

[0122] Compare the current structural offset with the structural offset of the previous period: if the offset increases, it indicates that the particle size distribution structure has further deviated from the normal distribution and further intervention is required; otherwise, it is judged that it will not deteriorate for the time being and the current state will be maintained.

[0123] Step 5: Execution of control logic

[0124] When the following conditions are met simultaneously: the disturbance level has not decreased for two consecutive cycles; and the structural offset shows an upward trend, the system will execute the following control strategy:

[0125] Keep the frequency of the crushing motor constant to avoid secondary variations in particle size distribution caused by frequency interference; reduce the screening speed to decrease the selective output tendency of particle size structure during screening and alleviate the accumulation of structural deviation.

[0126] Step Six: Monitor the width expansion ratio of the particle size distribution curve in real time.

[0127] In each sampling cycle, the system calculates the width expansion ratio of the new particle size distribution curve and compares it with the previous cycle to determine whether to continue expanding. If the width expansion ratio continues to increase in two consecutive cycles, the system continues to reduce the screening speed. If the expansion ratio stops increasing or decreases in a certain cycle, it can be determined that the structural expansion trend has been curbed, the system stops further reducing the speed, and records the current screening speed as a reference stable value.

[0128] In some embodiments of this application, when the control is determined to be invalid, the method further includes:

[0129] Analyze the changes in particle size distribution curves over two consecutive sampling cycles after adjusting the screening speed. When an increase in structural offset is detected, determine whether the disturbance source is a fluctuation in powder load based on the following conditions:

[0130] If, within two consecutive sampling cycles after the screening speed is adjusted, the dust concentration changes by more than 10% of the previous sampling, and the number of times this happens exceeds the preset number, it is determined that the concentration fluctuates frequently.

[0131] If, within two consecutive sampling periods after the screening speed is adjusted, the absolute value of the maximum change in particle size variation rate is greater than the baseline variation rate, it is determined that the particle size variation rate exhibits oscillation characteristics.

[0132] When the concentration fluctuates frequently and exhibits oscillation characteristics, the current disturbance is determined to be a fluctuation in powder load, and the drive frequency of the crushing motor and the screening speed are adjusted down simultaneously.

[0133] It should be noted that when the system is determined to be in an ineffective control state after performing one control operation, the following steps are further performed to identify whether the disturbance source belongs to powder load fluctuation, and control response is executed accordingly:

[0134] Step 1: Extract the detection cycle data after adjustment

[0135] The system selects two consecutive acquisition cycles after the screening speed is adjusted, and extracts the following three types of data as the basis for judgment: the change range of dust concentration in each sampling; the particle size variation rate corresponding to each sampling; and the structural offset of each particle size range in the current two cycles, which are used to judge the overall offset trend.

[0136] Step 2: Determine if the structural offset has increased.

[0137] Based on the particle size distribution curve data of all sampling points in two periods, the difference between the average particle size of each particle size interval and the standard reference particle size is calculated sequentially, and the sum is obtained to obtain the structural offset.

[0138] If the structural offset in the second cycle is higher than that in the first cycle, the system records it as "structural offset increased" and proceeds to the next judgment process; if it does not increase, the system will not proceed to the powder load disturbance judgment path, and will maintain the current parameter settings and continue monitoring.

[0139] Step 3: Determine if concentration fluctuations are frequent.

[0140] Given the assumption that the structural offset is increasing, the system performs frequency analysis on the dust concentration changes within two collection cycles: comparing the change range of the concentration value in each sample with that in the previous sample; and counting the number of times the change range exceeds 10% of the previous sample value.

[0141] If the number of occurrences exceeds the system's set reference number (e.g., three times), the system determines it as "frequent concentration fluctuations"; if it does not exceed the reference number, it will not proceed to the next step of judgment and will maintain the current control status.

[0142] Step 4: Determine if the particle size variation rate exhibits oscillatory characteristics.

[0143] Based on the frequent concentration fluctuations, the system further judges the fluctuation behavior of particle size variation rate during the sampling period: calculate the change range of particle size variation rate between each sampling point; extract the value with the largest change range and record it as the maximum change value; compare this value with the system's preset "normal fluctuation reference value"; if the maximum value is higher than the upper limit threshold of the normal reference value (e.g., 30%), it is judged that "the particle size variation rate has oscillation characteristics".

[0144] Step 5: Disturbance Source Identification and Control Triggering

[0145] When the following two conditions are met: frequent concentration fluctuations and oscillating characteristics in particle size variation rate, the system determines that the current disturbance source is "powder load fluctuation." At this time, the system immediately triggers a linkage control action: synchronously reducing the drive frequency of the crushing motor to reduce the frequency of raw material entry and particle impact, thereby mitigating the disturbance of released micro-powder; synchronously reducing the screening speed to slow down the pace of particle size structure adjustment and balance the backward distribution of fluctuations; and using the control moment as the starting point for a new round of evaluation to enter the next stage of particle size distribution monitoring.

[0146] In some embodiments of this application, after synchronously reducing the drive frequency of the crushing motor and the screening speed, the particle size distribution curve characteristics of subsequent sampling cycles are monitored.

[0147] If, in any period, the main peak of the particle size distribution curve moves toward the main peak of the standard particle size distribution curve, and the number of times the dust concentration change exceeds 10% of the previous sampling is less than or equal to three, it is considered an effective control.

[0148] If the main peak does not shift back, determine whether there is a particle size accumulation region. When the structural displacement in a certain particle size range is greater than the preset value of the structural displacement in the adjacent particle size range, and there is no main peak in the current particle size range, then there is a particle size accumulation region. Maintain the current crushing motor drive frequency and screening speed unchanged until the particle size accumulation tends to disappear.

[0149] In some embodiments of this application, after simultaneously reducing the drive frequency of the crushing motor and the screening speed, the system enters a continuous monitoring phase in subsequent sampling cycles to determine whether the control measures have produced a repair effect on the particle size structure. The monitoring and determination process includes:

[0150] First, the position of the main peak of the particle size distribution curve in each subsequent sampling period is analyzed and compared with the position of the main peak in the standard particle size distribution curve. When the main peak position of the current period is detected to have moved towards the standard main peak direction compared to before the control, and the number of times the dust concentration change exceeds 10% of the previous sampling period does not exceed three times, the system records this control as effective control, and considers that the particle size structure has tended to stabilize.

[0151] If no significant shift in the main peak position is observed within several consecutive cycles, further local analysis of the interval structure of the current particle size distribution curve is performed. Specifically, the current curve is divided into several particle size intervals, and the offset of the particle size distribution relative to the reference structure in each interval is calculated. When the offset of a certain interval is significantly higher than that of the adjacent particle size intervals, and the interval does not contain the main peak of the particle size distribution, it is determined that the interval has formed a particle size accumulation region. Such accumulation usually reflects a mismatch between the crushing output and the screening separation efficiency.

[0152] Upon confirming the existence of a particle size accumulation region, the system does not immediately change the current pulverizing motor drive frequency and screening speed. Instead, it enters a maintenance strategy phase, maintaining stable operation of the existing process parameters and continuing to track subsequent changes in particle size distribution. Only when the structural displacement amplitude in the particle size accumulation region significantly decreases and the distribution becomes more uniform does the system consider the accumulation phenomenon to have disappeared and end this phase of stable operation.

[0153] See Figure 2 As shown, this embodiment of the invention provides a quality control method for flour production, including:

[0154] S1: Collect the particle size distribution and concentration changes of airborne dust particles during flour production by deploying several aerosol particle detectors in the workshop;

[0155] S2: Calculate the disturbance level under the current process parameters based on the fluctuation of particle size distribution and concentration changes;

[0156] S3: Calculate the quality fluctuation value of the current batch of flour under the production status based on the disturbance level and particle size distribution;

[0157] S4: Based on the comparison results between the quality fluctuation value and the preset threshold, determine the mode switching and eliminate flour particle size fluctuation.

[0158] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A quality control system for flour production, characterized in that, include: The data acquisition module includes several aerosol particle detectors deployed in the workshop. The data acquisition module is configured to collect the particle size distribution and concentration changes of dust particles in the air during flour production. The disturbance level determination module, connected to the data acquisition module, is configured to calculate the disturbance level under the current process parameters based on the fluctuation degree of particle size distribution and concentration change. The quality fluctuation determination module, connected to the disturbance level determination module, is configured to calculate the quality fluctuation value under the current process parameters based on the disturbance level and particle size distribution. The monitoring mode switching module is configured to make a mode switching judgment based on the comparison result of the quality fluctuation value and the preset threshold. The mode switching judgment includes: when the quality fluctuation value is lower than the first threshold, maintaining the current working mode; when it is equal to or higher than the first threshold and lower than the second threshold, entering the enhanced monitoring mode, which includes adjusting the acquisition frequency according to the relationship between the quality fluctuation value and the second threshold; and when it is equal to or higher than the second threshold, entering the control mode to eliminate flour particle size fluctuation. The disturbance level is obtained by weighted summation based on the standard deviation of particle size distribution and concentration variation; during weighted summation, the contribution value calculated by the SHAP model is set as the weight coefficient. The steps for obtaining the quality fluctuation value include: Collect the disturbance level data sequence output by the disturbance level determination module within the current sampling period, denoted as D={d1,d2,…,dn}, where each disturbance level value dn, n is a positive integer, is calculated by weighting the standard deviation of particle size distribution and concentration change; assign the concentration change value to 0, construct the disturbance level standard sequence based on the standard particle size distribution of dust particles and the concentration change assigned to zero, obtain the absolute value of the difference between the disturbance level value in the disturbance level sequence at each time point within the sampling period and the standard disturbance level value corresponding to the disturbance level standard sequence, take the maximum value, and record it as the disturbance fluctuation amplitude; A particle size distribution curve is established based on the particle size distribution within the collection period. The particle size distribution curve is then divided into several particle size intervals. The difference between the mean actual distribution frequency and the reference value in each particle size interval is calculated, and the summation yields the structural offset. The mass fluctuation value is calculated based on the disturbance fluctuation amplitude and structural offset in the following form: ； Where Q represents the final mass fluctuation value, A represents the disturbance fluctuation amplitude, and S represents the structural offset; The sampling frequency in the enhanced monitoring mode is determined by the following method: The maximum sampling frequency of the aerosol particle detector is obtained, and the sampling frequency is (mass fluctuation value / second threshold) × maximum sampling frequency.

2. The quality control system for flour production according to claim 1, characterized in that, The monitoring mode switching module also includes a disturbance factor tracing unit, which is configured to, after the system enters the control mode: By comparing the trend of the main peak position of the particle size distribution curve and the degree of width expansion on both sides of the particle size distribution curve at the main peak position within several consecutive collection cycles, it is determined whether there is a phenomenon of particle size centroid shift and particle size structure distribution asymmetry. Adjustments are made based on the judgment results. If the expansion degree on either side is significantly higher than that on the other side and exceeds the set asymmetry threshold range, it can be judged as particle size structure distribution asymmetry. When both phenomena occur simultaneously, reduce the drive frequency of the crushing motor and the screening speed. When only particle size centroid shift exists, reduce the drive frequency of the crushing motor; When only the particle size distribution is asymmetrical, reduce the screening speed.

3. The quality control system for flour production according to claim 2, characterized in that, The monitoring mode switching module also includes a disturbance trend monitoring unit, which is configured as follows: After the disturbance factor tracing unit ends the regulation, the particle size distribution curve of the corresponding acquisition period at the time of regulation is analyzed. If the main peak position of the particle size distribution curve shifts back and the width expansion ratio of the particle size distribution curve is adjusted back, it is recorded as effective regulation; otherwise, it is judged as ineffective regulation. The expansion ratio is the sum of the interval widths of the distribution on both sides of the main peak particle size divided by the main peak intensity value. If the expansion ratio of the particle size distribution curve after regulation decreases compared with that before regulation, it is considered as a width expansion ratio adjustment; otherwise, it is judged as no adjustment.

4. The quality control system for flour production according to claim 3, characterized in that, When the control is deemed ineffective, the disturbance level and structural offset are obtained. If the disturbance level does not decrease within two consecutive acquisition cycles and the structural offset increases, the frequency of the crushing motor is kept constant and the screening speed is further reduced until the expansion ratio of the particle size distribution curve stops increasing.

5. The quality control system for flour production according to claim 4, characterized in that, When a regulation is deemed ineffective, it also includes: Analyze the changes in particle size distribution curves over two consecutive sampling cycles after adjusting the screening speed. When an increase in structural offset is detected, determine whether the disturbance source is a fluctuation in powder load based on the following conditions: If, within two consecutive sampling cycles after the screening speed is adjusted, the dust concentration changes by more than 10% of the previous sampling, and the number of times this happens exceeds a preset number, it is determined that the concentration fluctuates frequently. Within two consecutive sampling periods after the screening speed is adjusted, the variation value of particle size variation rate at each sampling point is calculated. When the absolute value of the maximum variation value of particle size variation rate is greater than the baseline variation rate, it is determined that the particle size variation rate has oscillation characteristics. When the concentration fluctuates frequently and exhibits oscillation characteristics, the current disturbance is determined to be a fluctuation in powder load, and the drive frequency of the crushing motor and the screening speed are adjusted down simultaneously.

6. The quality control system for flour production according to claim 5, characterized in that, After simultaneously reducing the drive frequency of the crushing motor and the screening speed, the particle size distribution curve characteristics of subsequent sampling cycles were monitored. If, in any period, the main peak of the particle size distribution curve moves toward the main peak of the standard particle size distribution curve, and the number of times the dust concentration change exceeds 10% of the previous sampling is less than or equal to three, it is considered an effective control. If the main peak does not shift back, determine whether there is a particle size accumulation region. Divide the current particle size distribution curve into several particle size intervals and calculate the offset of the particle size distribution relative to the reference structure in each particle size interval. When the offset of a certain interval is higher than that of the adjacent particle size interval, and there is no main peak in the current particle size interval, then there is a particle size accumulation region. Keep the current crushing motor drive frequency and screening speed unchanged until the particle size accumulation tends to disappear.

7. A quality control method for flour production, used to implement the quality control system for flour production as described in any one of claims 1-6, characterized in that, The method includes: S1: Collect the particle size distribution and concentration changes of airborne dust particles during flour production by deploying several aerosol particle detectors in the workshop; S2: Calculate the disturbance level under the current process parameters based on the fluctuation of particle size distribution and concentration changes; S3: Calculate the quality fluctuation value under the current process parameters based on the disturbance level and particle size distribution; S4: Based on the comparison result between the quality fluctuation value and the preset threshold, perform mode switching judgment; and eliminate flour particle size fluctuation.

Images