Method for monitoring quality stability for pesticide preparation and sensing system

Abstract

The application discloses a quality stability monitoring method and a sensing system for pesticide preparation, and belongs to the technical field of pesticide preparation quality monitoring. The method is arranged along a preparation pipeline in a sensing monitoring system of a pesticide preparation production line, and a plurality of sensing collection nodes are arranged. In the preparation process of a target pesticide preparation, sound and vibration response signals and medium density response signals of each monitoring section are collected. Disturbance propagation timing characteristics and density distribution gradient characteristics are extracted, section response stability is determined, and candidate response offset sections are marked. The section response stability is weighted and corrected based on the upstream and downstream correlation of the candidate response offset sections in the preparation pipeline, the corrected stability is obtained, and the quality stability state of the target pesticide preparation in the preparation process is determined. The scheme can improve the accuracy of local quality offset identification and reduce the misjudgment caused by single-section instantaneous disturbance.

Description

Technical Field

[0001] This invention relates to the field of pesticide preparation quality monitoring technology, specifically to a method and sensing system for monitoring the quality stability of pesticide preparation. Background Technology

[0002] The pesticide formulation preparation process involves the continuous transport, mixing, and dispersion of multiple components in the preparation pipeline. The flow state, density distribution, and local disturbances of the materials in different sections of the pipeline directly affect the final formulation quality. Current quality stability monitoring methods mostly employ offline sampling analysis or online measurement of a single parameter, such as installing temperature, pressure, flow, or density sensors in local areas of the pipeline, and using threshold values ​​to determine whether deviations have occurred in the preparation process. Some online monitoring schemes collect acoustic or density signals and use overall signal amplitude, mean, or spectral changes as the basis for stability judgment. However, anomalies such as uneven mixing, phase separation, decreased dispersion stability, or local agglomeration in the formulation within the preparation pipeline often manifest as segmental disturbance propagation changes and density distribution gradient changes. Conventional single-parameter or overall statistical methods are insufficient to distinguish between normal process fluctuations and local quality deviations, and cannot locate abnormal sections in continuous pipelines. Since there is continuous material flow and disturbance transmission between upstream and downstream sections of the preparation pipeline, instantaneous fluctuations in a local section may be caused by short-term disturbances upstream and do not necessarily indicate an abnormality in the quality stability of that section itself. Existing monitoring methods do not adequately utilize the correlations between different segments, easily misjudging transient disturbances in a single segment as quality anomalies, leading to misjudgments or tracing errors. How to obtain segment-level process information during preparation and reduce the impact of transient disturbances in a single segment on the judgment results under continuous multi-segment monitoring conditions is a problem that needs to be solved in the current monitoring of pesticide preparation quality stability. Summary of the Invention

[0003] To address the challenges of existing technologies in locating localized mass shifts in continuous preparation pipelines and the potential for misjudgments due to localized transient disturbances, this paper provides a quality stability monitoring method and sensing system for pesticide preparation. The system determines the segmental response stability of each monitoring segment using acoustic vibration response signals and medium density response signals, performs upstream and downstream correlation weighted correction on candidate response shift segments, and determines the quality stability status of the target pesticide formulation during preparation.

[0004] To achieve the above objectives, the present invention provides the following technical solution: The present invention provides a method and sensing system for monitoring the quality stability of pesticide preparation. The method is applied to a sensing monitoring system in a pesticide preparation production line. The sensing monitoring system includes multiple sets of sensing acquisition nodes arranged sequentially along the preparation pipeline. The method includes: during the preparation of the target pesticide formulation, acquiring acoustic vibration response signals and medium density response signals of corresponding monitoring sections in the preparation pipeline through each of the sensing acquisition nodes; extracting the disturbance propagation time sequence characteristics of each monitoring section based on the acoustic vibration response signals of each monitoring section, and extracting the density distribution gradient characteristics of each monitoring section based on the medium density response signals of each monitoring section; and further extracting the disturbance propagation time sequence characteristics of each monitoring section based on the medium density response signals of each monitoring section. The disturbance propagation time sequence characteristics and density distribution gradient characteristics of the monitoring sections are used to determine the segment response stability of each monitoring section. Based on the segment response stability of each monitoring section, candidate response offset sections are marked within each monitoring section. Based on the upstream and downstream correlation of the candidate response offset sections in the preparation pipeline, the segment response stability of the candidate response offset sections is weighted and corrected to obtain the corrected stability of each candidate response offset section. The quality stability status of the target pesticide formulation during the preparation process is determined based on the corrected stability. By jointly analyzing the acoustic vibration response and the medium density response, and using the upstream and downstream correlation to correct the segment stability, quality stability anomalies during the preparation process can be identified more accurately, reducing misjudgments caused by single-point signal fluctuations.

[0005] As a technical solution of the present invention, when acquiring acoustic vibration response signals and medium density response signals of corresponding monitoring sections in the preparation pipeline through the sensing acquisition nodes, the acoustic vibration sensors in each sensing acquisition node acquire the original acoustic vibration signals of the corresponding monitoring sections at a first sampling frequency, and perform frequency band separation processing on the original acoustic vibration signals to obtain high-frequency acoustic vibration components and low-frequency acoustic vibration components; the density sensing units in each sensing acquisition node acquire the original density signals of the corresponding monitoring sections at a second sampling frequency, where the second sampling frequency is lower than the first sampling frequency; the high-frequency acoustic vibration components, the low-frequency acoustic vibration components, and the original density signals are aligned according to the acquisition time and combined to form the acoustic vibration response signals and medium density response signals of the corresponding monitoring sections. By synchronously acquiring and time-aligning the high-frequency acoustic vibration components, low-frequency acoustic vibration components, and density signals, the time correspondence between different response signals can be preserved, improving the accuracy of subsequent feature extraction.

[0006] Preferably, when extracting the disturbance propagation timing features of each monitoring segment based on the acoustic-vibration response signal of each monitoring segment, for any given monitoring segment, the peak amplitude sequence of the high-frequency acoustic-vibration component within multiple acquisition windows and the zero-crossing rate sequence of the low-frequency acoustic-vibration component within the same multiple acquisition windows are extracted from the acoustic-vibration response signal of that monitoring segment; the cross-correlation offset between the peak amplitude sequence and the zero-crossing rate sequence on the time axis is calculated; the disturbance propagation delay of the monitoring segment is determined based on the cross-correlation offset, and this disturbance propagation delay is used as the disturbance propagation timing feature of the monitoring segment. This disturbance propagation timing feature can reflect the time delay of the acoustic-vibration response within the monitoring segment, which helps to capture the temporal changes in disturbance propagation.

[0007] Preferably, when extracting the density distribution gradient features of each monitoring segment based on the medium density response signal of each monitoring segment, for any given monitoring segment, the medium density response signal of that monitoring segment is divided into multiple density sampling sequences according to the acquisition time, and each density sampling sequence corresponds to an acquisition period; for any given density sampling sequence, the density change slope between adjacent sampling points in the density sampling sequence is calculated, and all the calculated density change slopes are arranged in chronological order to obtain the slope change trajectory of that acquisition period; the slope span value between the maximum and minimum slope values ​​in the slope change trajectory is calculated, and the slope span value of that acquisition period is used as the density distribution gradient feature of the monitoring segment in that acquisition period. This density distribution gradient feature can quantify the gradient change range of the medium density within the acquisition period, providing a basis for identifying density distribution anomalies.

[0008] Preferably, when determining the segment response stability of each monitoring segment based on the disturbance propagation time-series characteristics and density distribution gradient characteristics of each monitoring segment, for any given monitoring segment, the disturbance delay fluctuation amplitude of that monitoring segment over multiple consecutive acquisition cycles is determined based on the disturbance propagation time-series characteristics of that monitoring segment; the density gradient change amplitude of that monitoring segment over the same multiple consecutive acquisition cycles is determined based on the density distribution gradient characteristics of that monitoring segment; the disturbance delay fluctuation amplitude and the density gradient change amplitude are then fused according to a preset weight to obtain the segment response stability of that monitoring segment over the same multiple consecutive acquisition cycles. The segment response stability is negatively correlated with both the disturbance delay fluctuation amplitude and the density gradient change amplitude. This unifies the disturbance delay fluctuation and density gradient change into a single stability metric, facilitating horizontal comparisons across multiple monitoring segments.

[0009] Preferably, when marking candidate response offset segments in each monitoring segment based on the segment response stability of each monitoring segment, the segment response stability of each monitoring segment is arranged into a stability sequence according to the arrangement order of each monitoring segment in the preparation pipeline; a local minimum point is determined in the stability sequence, wherein the segment response stability of the local minimum point is less than the response stability of its two adjacent segments in the stability sequence; when the segment response stability of the local minimum point is lower than a preset stability lower limit threshold, the monitoring segment corresponding to the local minimum point is marked as a candidate response offset segment. Through the dual screening of local minimum points and the stability lower limit threshold, the mislabeling of candidate segments caused by random fluctuations can be reduced.

[0010] Preferably, based on the upstream and downstream correlation of the candidate response offset segments in the preparation pipeline, when weighted correction is applied to the segment response stability of the candidate response offset segments to obtain the corrected stability of each candidate response offset segment, for any candidate response offset segment, the upstream adjacent monitoring segment and the downstream adjacent monitoring segment are determined; the first segment response stability of the upstream adjacent monitoring segment and the second segment response stability of the downstream adjacent monitoring segment are obtained; based on the degree of difference between the first segment response stability and the second segment response stability, a target associated segment with a correlation fluctuation trend with the candidate response offset segment is determined among the upstream adjacent monitoring segment and the downstream adjacent monitoring segment; the segment response stability of the candidate response offset segment is weighted and corrected based on the segment response stability of the target associated segment to obtain the corrected stability of the candidate response offset segment. Correcting the candidate segment by the correlation fluctuation trend of adjacent segments can more realistically reflect whether the candidate segment belongs to a local offset within the overall trend of the pipeline.

[0011] Furthermore, based on the degree of difference between the response stability of the first segment and the response stability of the second segment, when determining the target associated segment among the upstream adjacent monitoring segment and the downstream adjacent monitoring segment that has a related fluctuation trend with the candidate response offset segment, the absolute difference between the response stability of the first segment and the response stability of the second segment is calculated; when the absolute difference is less than a preset difference threshold, both the upstream adjacent monitoring segment and the downstream adjacent monitoring segment are taken as target associated segments; when the absolute difference is greater than or equal to the preset difference threshold, the monitoring segment with the smaller absolute difference between the response stability of the first segment and the response stability of the second segment and the segment response stability of the candidate response offset segment is selected. The test section is designated as the target associated section. When obtaining the corrected stability of the candidate response offset section, the pipeline distance between the candidate response offset section and its target associated section is acquired, and a stability weighting coefficient for the target associated section is determined based on this pipeline distance. This stability weighting coefficient is negatively correlated with the pipeline distance. The segment response stability of the candidate response offset section and the segment response stability of the target associated section are weighted and summed according to the stability weighting coefficient to obtain a weighted stability. A trend correction factor is determined based on the changing trend of the segment response stability of multiple monitoring sections upstream of the candidate response offset section. The weighted stability is then corrected using the trend correction factor to obtain the corrected stability of the candidate response offset section. By combining the pipeline distance weighting and the upstream changing trend for correction, the adaptability and accuracy of the corrected stability to the actual process can be improved.

[0012] Preferably, when determining the quality stability status of the target pesticide formulation during the preparation process based on the corrected stability, the corrected stability of each candidate response offset segment is compared with a preset corrected stability threshold. When the corrected stability of all candidate response offset segments is greater than or equal to the corrected stability threshold, the quality stability status of the target pesticide formulation during the preparation process is determined to be normal. When the corrected stability of at least one candidate response offset segment is less than the corrected stability threshold, the quality stability status of the target pesticide formulation during the preparation process is determined to be abnormal, and the candidate response offset segments with corrected stability less than the corrected stability threshold are marked as quality offset traceability segments. This enables the determination of quality stability anomalies and segment-level traceability, providing location information for quality control in the pesticide preparation process.

[0013] This invention also provides a quality stability monitoring sensing system for pesticide preparation, including a memory, a processor, and a computer program stored in the memory and running on the processor. When executing the computer program, the processor implements the steps of the quality stability monitoring method for pesticide preparation provided by this invention. This sensing system can process acoustic and vibration response signals and medium density response signals in the preparation pipeline online, automatically completing stability analysis, candidate section correction, and quality stability status determination.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0015] Using acoustic vibration response signals and medium density response signals as the fusion information sources for section stability determination, the peak amplitude sequence and zero-crossing rate sequence of the acoustic vibration response signal are extracted separately for the high-frequency and low-frequency bands, and the disturbance propagation delay is obtained by the cross-correlation offset of the two on the time axis. The medium density response signal is divided into density sampling sequences according to the acquisition period, and the density change slope between adjacent sampling points is calculated, with the slope span value as the density distribution gradient feature. The disturbance delay fluctuation amplitude and density gradient change amplitude within multiple consecutive acquisition periods are fused according to preset weights to obtain the section response stability. This method incorporates the propagation delay change of mechanical disturbances in the pipeline and the distribution change of medium density along the time direction into the same section measurement, no longer relying on a single density mean or overall acoustic vibration amplitude for stability judgment. It has a more direct response capability to disturbance transmission anomalies and density distribution anomalies caused by local mixing inhomogeneity, phase separation, or decreased dispersion stability, and can refine the quality offset identification to specific monitoring sections, reducing the masking effect of normal process fluctuations on the identification of local anomalies. When weighted correcting candidate response offset segments, the segment response stability of the upstream and downstream adjacent monitoring segments is obtained. Based on the absolute difference between the upstream and downstream stability, a target associated segment with a related fluctuation trend is determined, and a stability weight coefficient is determined based on pipeline distance, with greater weight for closer pipelines. The segment response stability of the candidate segment and the target associated segment are weighted and summed, then multiplied by a trend correction factor obtained from the stability change trends of multiple upstream segments to form the corrected stability of the candidate segment. This upstream-downstream correlation weighted correction allows the stability determination of candidate segments to be verified within the continuous flow context of the preparation pipeline. Instantaneous disturbances in a single segment will not directly trigger quality anomaly determination in the absence of a correlation trend between adjacent segments. The true quality offset can be distinguished from upstream transitive disturbances, improving the accuracy and reliability of quality offset tracing segment marking. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is a flowchart of a quality stability monitoring method for pesticide preparation;

[0018] Figure 2 This is a flowchart of the acoustic vibration response signal and the medium density response signal processing and feature extraction.

[0019] Figure 3 This is a schematic diagram illustrating the timing characteristics of the acoustic vibration response signal and the medium density response signal in the monitoring section;

[0020] Figure 4 These are the curves showing the change in response stability and corrected stability of the candidate response offset segment. Detailed Implementation

[0021] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, 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.

[0022] See Figure 1 This invention provides a method for monitoring the quality stability of pesticide preparation, applied to a sensor monitoring system in a pesticide preparation production line. The sensor monitoring system includes multiple sets of sensor acquisition nodes arranged sequentially along the preparation pipeline. During the preparation of the target pesticide formulation, acoustic vibration response signals and medium density response signals of corresponding monitoring sections in the preparation pipeline are acquired through each sensor acquisition node. The disturbance propagation time-series characteristics of each monitoring section are extracted based on the acoustic vibration response signals, and the density distribution gradient characteristics of each monitoring section are extracted based on the medium density response signals. The segment response stability of each monitoring section is determined based on the disturbance propagation time-series characteristics and density distribution gradient characteristics. Candidate response deviation sections are marked within each monitoring section based on their segment response stability. The segment response stability of the candidate response deviation sections is weighted and corrected based on their upstream and downstream correlation in the preparation pipeline to obtain the corrected stability of each candidate response deviation section. The quality stability status of the target pesticide formulation during the preparation process is determined based on the corrected stability.

[0023] In specific implementation, please refer to Figure 2 When acquiring acoustic vibration response signals and medium density response signals of the corresponding monitoring sections in the preparation pipeline through various sensing nodes, the acoustic vibration sensors in each sensing node acquire the raw acoustic vibration signals of the corresponding monitoring sections at a first sampling frequency. The first sampling frequency is determined based on the effective response bandwidth of the acoustic vibration sensor, and is set to be more than twice the upper limit frequency of the effective response bandwidth of the acoustic vibration sensor. The upper limit frequency of the effective response bandwidth of the acoustic vibration sensor is 5000Hz, and the first sampling frequency is set to 20000Hz. When performing frequency band separation processing on the raw acoustic vibration signals, a bandpass filter is used to perform frequency band separation processing on the raw acoustic vibration signals. The bandpass filter includes a high-pass filter and a low-pass filter. The cutoff frequency of the high-pass filter is set to 500Hz to obtain high-frequency acoustic vibration components with frequencies greater than 500Hz, and low-frequency acoustic vibration components correspond to acoustic vibration signal components with frequencies less than or equal to 500Hz. The cutoff frequency of 500Hz is determined based on the distribution characteristics of acoustic vibration energy generated by fluid disturbance in the preparation pipeline. The acoustic vibration energy generated by fluid disturbance in the preparation pipeline is in the range of 10Hz to 5000Hz. Among them, the acoustic vibration component with a frequency greater than 500Hz is related to the collision of local bubbles or particles in the preparation pipeline, and the acoustic vibration component with a frequency less than or equal to 500Hz is related to the overall fluctuation of the fluid medium in the preparation pipeline.

[0024] The original density signal of the corresponding monitoring segment is acquired by the density sensing unit in each sensing node at a second sampling frequency. The second sampling frequency is lower than the first sampling frequency. The second sampling frequency is determined based on the response time of the density sensing unit, which is 0.1 seconds, and the second sampling frequency is set to 10Hz. After aligning the high-frequency acoustic vibration component, the low-frequency acoustic vibration component, and the original density signal according to the acquisition time, they are combined into the acoustic vibration response signal and the medium density response signal of the corresponding monitoring segment. During alignment, the original density signal is linearly interpolated on the reference time axis using the acquisition time of the acoustic vibration sensor as the reference time axis to obtain the original density signal corresponding to the same acquisition time of the high-frequency and low-frequency acoustic vibration components. The combined acoustic vibration response signal consists of the high-frequency and low-frequency acoustic vibration components, and the medium density response signal is the original density signal.

[0025] When extracting the disturbance propagation time-series features of any monitoring segment based on its acoustic-vibration response signal, the peak amplitude sequence of the high-frequency acoustic-vibration component within multiple acquisition windows and the zero-crossing rate sequence of the low-frequency acoustic-vibration component within multiple acquisition windows are extracted from the acoustic-vibration response signal of that monitoring segment. The acquisition window length is set to 100ms, and the step size between adjacent acquisition windows is set to 50ms. The acquisition window length is set based on the fact that the time scale of the disturbance propagation delay is on the order of seconds. A 100ms acquisition window length can retain complete disturbance peak information within the acquisition window and avoid the time-series resolution being reduced due to an excessively large acquisition window length. The step size is set to half the acquisition window length to retain the overlapping part between adjacent acquisition windows and improve the continuity of time-series feature extraction.

[0026] For each acquisition window, the amplitude with the largest absolute value in the high-frequency acoustic vibration component is taken as the peak amplitude of that acquisition window. The peak amplitudes of multiple acquisition windows are arranged in chronological order according to the acquisition windows to obtain a peak amplitude sequence. For each acquisition window, the number of times the signal in the low-frequency acoustic vibration component crosses zero is divided by the acquisition window length to obtain the zero-crossing rate of that acquisition window. The zero-crossing rates of multiple acquisition windows are arranged in chronological order according to the acquisition windows to obtain a zero-crossing rate sequence.

[0027] When calculating the cross-correlation offset between the peak amplitude sequence and the zero-crossing rate sequence on the time axis, the peak amplitude sequence and the zero-crossing rate sequence are multiplied point-by-point and summed under different lag window numbers to obtain the cross-correlation values ​​for different lag window numbers. The lag window number ranges from the negative maximum lag window number to the positive maximum lag window number. The maximum lag window number is set to one-quarter of the total number of acquisition windows. The basis for setting the maximum lag window number to one-quarter of the total number of acquisition windows is that the disturbance propagation delay does not exceed one-quarter of the total duration of the acoustic vibration response signal.

[0028] The maximum hysteresis window value among the cross-correlation values ​​under different hysteresis window numbers is used as the cross-correlation offset. When determining the disturbance propagation delay for this monitoring segment based on the cross-correlation offset, the cross-correlation offset is multiplied by the step size to obtain the disturbance propagation delay. The disturbance propagation delay is then used as the disturbance propagation time series characteristic of this monitoring segment.

[0029] When extracting the density distribution gradient features of any monitoring segment based on its density response signal, the signal is divided into multiple density sampling sequences according to the acquisition time. Each sampling sequence corresponds to one acquisition period. The acquisition period is set to 10 seconds, and the step size between adjacent acquisition periods is set to 5 seconds. The 10-second acquisition period is chosen because the density fluctuation period in the density response signal ranges from several seconds to tens of seconds, and a 10-second acquisition period can cover at least one complete density fluctuation process. The step size is set to half the acquisition period length to preserve the overlap between adjacent acquisition periods and improve the continuity of density distribution gradient feature extraction.

[0030] For any density sampling sequence, calculate the slope of the density change between adjacent sampling points in the sequence. The density change slope is calculated using the following formula:

[0031]

[0032] in, Indicates the first The density sampling point and the first The slope of density change between density sampling points Indicates the first The density value of the medium at each density sampling point Indicates the first The density value of the medium at each density sampling point Indicates the first The acquisition time of each density sampling point Indicates the first The acquisition time of each density sampling point.

[0033] Arrange all calculated density change slopes in chronological order to obtain the slope change trajectory for that acquisition period. Calculate the slope span between the maximum and minimum slope values ​​in the slope change trajectory; the slope span value is the difference between the maximum and minimum slope values. Use the slope span value for that acquisition period as the density distribution gradient characteristic of that monitoring segment during that acquisition period.

[0034] See Figure 3 In the graph, the horizontal axis represents the time range of 0 to 3 seconds, the left vertical axis represents the acoustic amplitude value, ranging from -1.5 to 2.5, and the right vertical axis represents the medium density value, in kilograms per cubic meter (kg / m³), ranging from 1000 to 1080. The curves and sequences include: solid blue lines representing high-frequency acoustic components, dashed orange lines representing low-frequency acoustic components, solid purple lines representing the original density signal, green dotted lines representing the peak amplitude sequence, and red dashed lines representing the zero-crossing rate sequence amplified 10 times.

[0035] As shown in the figure, the high-frequency acoustic vibration component exhibits multiple pulse peaks with amplitudes fluctuating between 0 and 1.7. The peak amplitude sequence largely corresponds to the pulse peaks of the high-frequency acoustic vibration component, demonstrating accurate extraction of the acoustic vibration signal peaks. The low-frequency acoustic vibration component curve shows continuous fluctuations with an amplitude range of approximately -0.5 to 0.5, reflecting the overall fluctuation characteristics of the fluid medium inside the preparation pipeline. The zero-crossing rate sequence is marked with a red dashed line, and after magnification, it shows multiple distinct zero-crossing points, reflecting the dynamic changes in the zero-crossing points of the low-frequency acoustic vibration signal. Together with the peak amplitude sequence, this constitutes the disturbance propagation timing characteristics of the monitoring section.

[0036] The original density signal is displayed as a purple curve, exhibiting a slow, fluctuating trend, with the medium density value fluctuating between 1028 and 1070 kg / m³. The relatively long period of this density signal reflects the dynamic distribution characteristics of the medium density. The high and low frequency components of the acoustic vibration signal are synchronously aligned with the medium density signal on the time axis, ensuring the temporal consistency between the acoustic vibration response signal and the medium density response signal, thus meeting the signal acquisition and processing requirements of the above embodiment.

[0037] In practice, for any given monitoring segment, the disturbance propagation time sequence characteristics of that segment are used to determine the disturbance delay fluctuation amplitude over multiple consecutive acquisition cycles. The disturbance propagation time sequence characteristics are defined as the disturbance propagation delay amount, with one disturbance propagation delay amount corresponding to each acquisition cycle. Multiple consecutive acquisition cycles are set to 10 acquisition cycles, consistent with the acquisition cycle length used when extracting the density distribution gradient characteristics. The 10 acquisition cycles are chosen to cover at least two complete disturbance fluctuation processes within the preparation pipeline, while avoiding excessively long time windows that could delay the quality stability response. The disturbance propagation delay amount corresponding to each of the 10 consecutive acquisition cycles is obtained, resulting in 10 disturbance propagation delay amounts. The difference between the maximum and minimum disturbance propagation delay amounts among the 10 amounts is calculated, and this difference is taken as the disturbance delay fluctuation amplitude.

[0038] Based on the density distribution gradient characteristics of the monitoring section, the magnitude of density gradient change in the monitoring section over multiple consecutive acquisition cycles is determined. The density distribution gradient characteristic is a slope span value, with one slope span value corresponding to each acquisition cycle. Multiple consecutive acquisition cycles are also set to 10 acquisition cycles, consistent with the acquisition cycle used for disturbance delay fluctuation amplitude. The slope span value corresponding to each of the 10 consecutive acquisition cycles is obtained, resulting in 10 slope span values. The difference between the maximum and minimum slope span values ​​among the 10 slope span values ​​is calculated, and this difference is taken as the density gradient change magnitude.

[0039] The disturbance delay fluctuation amplitude and density gradient change amplitude are fused according to preset weights to obtain the section response stability of the monitoring section over 10 consecutive acquisition cycles. The fusion calculation uses the following formula:

[0040]

[0041] in, Indicates the first in the preparation pipeline The stability of the section response over 10 consecutive data collection cycles for each monitoring section. To monitor the sequence number of the monitoring section in the preparation pipeline, Indicates the first The amplitude of disturbance delay fluctuation in each monitoring segment over 10 consecutive data collection periods. Indicates the first The magnitude of density gradient change in a monitoring section over 10 consecutive data collection periods. This represents the weighting coefficient for the fusion of disturbance delay fluctuation amplitude. The weighting coefficients represent the magnitude of the density gradient change. Set to 0.01, weighting coefficient Set to 0.02. Weighting coefficient. and weighting coefficients The settings are based on the following: In the historical operating data of the sensor monitoring system, the amplitude of disturbance delay fluctuations and the amplitude of density gradient changes are normalized. The typical range of disturbance delay fluctuation amplitude is 0 to 100 milliseconds, and the typical range of density gradient change amplitude is 0 to 50 kg / m³ / s. According to the principle that the mass stability state is more sensitive to density gradient change amplitude than to disturbance delay fluctuation amplitude, the fusion weighting coefficient is set. With fusion weight coefficient The ratio is 1:2, which makes the segment response stability twice as responsive to changes in the density gradient amplitude as it is to changes in the disturbance delay fluctuation amplitude. Therefore, it is set as follows: It is 0.01. It is 0.02.

[0042] The stability of the section response is negatively correlated with the amplitude of the disturbance delay fluctuation and the amplitude of the density gradient change. When the amplitude of the disturbance delay fluctuation or the amplitude of the density gradient change increases, the stability of the section response decreases.

[0043] When marking candidate response offset segments based on the segment response stability of each monitoring segment, the segment response stability of each monitoring segment is arranged into a stability sequence according to the order in which the monitoring segments are arranged in the preparation pipeline. The stability sequence is... The element is the first The segment response stability of each monitoring segment is determined. When identifying local minima in the stability sequence, for each element in the stability sequence except the first and last elements, the segment response stability corresponding to the current element is compared with the segment response stability corresponding to the preceding and following elements in the stability sequence. When the segment response stability corresponding to the current element is less than the segment response stability corresponding to the preceding and following elements, the monitoring segment corresponding to the current element is identified as a local minimum. When the segment response stability of the local minimum is lower than a preset stability lower limit threshold, the monitoring segment corresponding to the local minimum is marked as a candidate response offset segment. The stability lower limit threshold is set to 0.3. The basis for setting the stability lower limit threshold to 0.3 is that when the segment response stability is lower than 0.3, the disturbance delay fluctuation amplitude and density gradient change amplitude reach a significant abnormal level after fusion calculation. At this level, the response offset state of the monitoring segment needs to enter the subsequent weighted correction process.

[0044] In practice, for any candidate response offset segment, the upstream and downstream adjacent monitoring segments are determined. Each candidate response offset segment corresponds to a sensor node in the preparation pipeline. The upstream adjacent monitoring segment is the monitoring segment corresponding to the first sensor node along the material flow direction in the preparation pipeline that precedes the sensor node corresponding to the candidate response offset segment. The downstream adjacent monitoring segment is the monitoring segment corresponding to the first sensor node along the material flow direction in the preparation pipeline that follows the sensor node corresponding to the candidate response offset segment. The candidate response offset segment is marked by local minima in the stability sequence. Since the candidate response offset segment is neither the first nor the last element in the stability sequence, both upstream and downstream adjacent monitoring segments exist.

[0045] The system acquires the first segment response stability of the upstream adjacent monitoring segment and the second segment response stability of the downstream adjacent monitoring segment. The first segment response stability is the segment response stability of the upstream adjacent monitoring segment over 10 consecutive acquisition cycles, and the second segment response stability is the segment response stability of the downstream adjacent monitoring segment over 10 consecutive acquisition cycles. The first and second segment response stability values ​​are read from the segment response stability data of each monitoring segment stored in the sensor monitoring system.

[0046] Calculate the absolute difference between the response stability of the first segment and the response stability of the second segment. The preset difference threshold is set to 0.1. The basis for setting the difference threshold to 0.1 is that the segment response stability value is in the range of 0 to 1. When the absolute difference between the response stability of the first segment and the response stability of the second segment is less than 0.1, the segment response stability values ​​of the upstream adjacent monitoring segment and the downstream adjacent monitoring segment are relatively close, and they have a correlated fluctuation trend with the candidate response offset segment. When the absolute difference is greater than or equal to 0.1, only one of the upstream adjacent monitoring segment and the downstream adjacent monitoring segment is closer to the candidate response offset segment in terms of stability value.

[0047] When the absolute difference is less than a preset difference threshold, both the upstream and downstream adjacent monitoring segments are considered as target associated segments. When the absolute difference is greater than or equal to the preset difference threshold, the monitoring segment with the smaller absolute difference between the first segment response stability and the second segment response stability and the segment response stability of the candidate response offset segment is considered as the target associated segment. The segment response stability of the candidate response offset segment is read from the segment response stability data of each monitoring segment stored in the sensor monitoring system.

[0048] When obtaining the pipeline distance between the candidate response offset segment and the target associated segment, the pipeline distance is the actual installation path length of the sensor acquisition node corresponding to the candidate response offset segment and the sensor acquisition node corresponding to the target associated segment on the prepared pipeline. The sensor monitoring system stores the installation position coordinates of each sensor acquisition node. Based on the installation position coordinates of the sensor acquisition node corresponding to the candidate response offset segment and the sensor acquisition node corresponding to the target associated segment, the pipeline path length between the two nodes is calculated on the three-dimensional pipeline path model of the prepared pipeline to obtain the pipeline distance. When there are two target associated segments, the first pipeline distance between the candidate response offset segment and the upstream adjacent monitoring segment, and the second pipeline distance between the candidate response offset segment and the downstream adjacent monitoring segment are obtained respectively.

[0049] The stability weighting coefficient for the target associated section is determined based on the pipeline distance, and the stability weighting coefficient is negatively correlated with the pipeline distance. When there is only one target associated section, the stability weighting coefficient is calculated according to the following formula:

[0050]

[0051] in, This represents the stability weight coefficient corresponding to the target associated segment. Indicates the pipeline distance attenuation coefficient. This represents the pipeline distance between the candidate response offset segment and the target associated segment, expressed in meters. Pipeline distance attenuation coefficient. Set to 0.05 per meter. Pipeline distance attenuation coefficient. The basis for setting it to 0.05 per meter is that the typical pipeline distance between adjacent sensor acquisition nodes in the sensor monitoring system is in the range of 2 to 10 meters. When the pipeline distance is 2 meters, the stability weight coefficient is 0.909, and when the pipeline distance is 10 meters, the stability weight coefficient is 0.667. The larger the pipeline distance, the smaller the stability weight coefficient, which makes the correction effect of the segment response stability of the target associated segment on the candidate response offset segment weaken as the pipeline distance increases.

[0052] When there are two target associated sections, the first pipeline distance between the candidate response offset section and the upstream adjacent monitoring section is substituted into the above stability weight coefficient calculation formula to obtain the upstream stability weight coefficient; the second pipeline distance between the candidate response offset section and the downstream adjacent monitoring section is substituted into the above stability weight coefficient calculation formula to obtain the downstream stability weight coefficient. The average of the upstream and downstream stability weight coefficients is used as the stability weight coefficient of the target associated section.

[0053] The weighted stability of the candidate response offset segment is obtained by weighting and summing the segment response stability of the target associated segment with the segment response stability of the target associated segment according to a stability weighting coefficient. When there is only one target associated segment, the segment response stability of the target associated segment is taken as the segment response stability of that target associated segment. When there are two target associated segments, the segment response stability of the target associated segment is taken as the average of the first segment response stability of the upstream adjacent monitoring segment and the second segment response stability of the downstream adjacent monitoring segment.

[0054] Based on the changing trends of the segment response stability of multiple monitoring segments upstream of the candidate response offset segment, a trend correction factor for the candidate response offset segment is determined. Five consecutive monitoring segments upstream of the candidate response offset segment are selected, and the segment response stability of these five segments is obtained. These segments are arranged in order of distance from the candidate response offset segment, as follows: first upstream stability, second upstream stability, third upstream stability, fourth upstream stability, and fifth upstream stability. Least squares linear fitting is performed on the first, second, third, fourth, and fifth upstream stability segments to obtain the linear fitting slope. The linear fitting slope represents the changing trend of the segment response stability of multiple monitoring segments upstream. The linear fitting slope is limited to the range of -0.2 to 0.2. When the linear fitting slope is less than -0.2, the trend correction factor is set to -0.2; when the linear fitting slope is greater than 0.2, the trend correction factor is set to 0.2; when the linear fitting slope is within the range of -0.2 to 0.2, the linear fitting slope is directly used as the trend correction factor. The basis for setting the trend correction factor to a range of -0.2 to 0.2 is that the numerical range of segment response stability is 0 to 1, and the correction effect of the stability change trend in the upstream direction on the candidate response offset segment should not exceed 20% of the segment response stability of the candidate response offset segment itself.

[0055] Multiplying the weighted stability of the candidate response offset segment by the trend correction factor yields the corrected stability of that segment. The corrected stability is calculated using the following formula:

[0056]

[0057] in, Indicates the first Corrected stability of each candidate response offset segment The sequential numbering of the candidate response offset segment when it is marked. Indicates the first The segment response stability of each candidate response offset segment. Indicates the first The segment response stability of the target associated segment for each candidate response offset segment, when there are two target associated segments. The average value of the response stability of the first segment of the upstream adjacent monitoring section and the response stability of the second segment of the downstream adjacent monitoring section is taken. This represents the stability weight coefficient corresponding to the target associated segment. When there are two target associated segments... Take the average of the upstream stability weighting coefficient and the downstream stability weighting coefficient. Indicates the first The trend correction factor for each candidate response offset segment is obtained by least-squares linear fitting of the segment response stability of five consecutive monitoring segments in the upstream direction of the candidate response offset segment and is limited to the range of -0.2 to 0.2.

[0058] See Figure 4 In the figure, the horizontal axis represents the candidate response offset segment number, ranging from 1 to 100; the left side of the vertical axis represents the stability value, ranging from 0 to 1; and the right side of the vertical axis represents the trend correction factor, ranging from -0.2 to 0.2. The figure shows four curves: segment response stability, weighted stability, corrected stability, and trend correction factor.

[0059] The segment response stability curve is marked with blue dashed dots. The overall stability value fluctuates between 0.1 and 0.35. The value is low and the fluctuation range is large, reflecting that the stability of the candidate response offset segment is poor after preliminary fusion calculation.

[0060] The weighted stability curve is marked with an orange dotted-line square. The value is significantly higher than that of the segment response stability, mainly distributed in the range of 0.35 to 0.65. This indicates that after the stability weighting correction based on the upstream and downstream related segments, the stability of the candidate response offset segment is improved and the fluctuation tends to be smoother.

[0061] The corrected stability curve is marked with a solid green triangle. Its value is similar to the overall trend of the weighted stability, but it shows more obvious fluctuations in some sections. The value range is about 0.3 to 0.7, which reflects the further adjustment effect of the trend correction factor on the weighted stability.

[0062] The trend correction factor curve is marked with a red dotted diamond shape and distributed on the right side of the vertical axis. The value fluctuates between -0.2 and 0.2, showing the linear fitting slope of the stability change trend of the five monitoring segments upstream of the candidate response offset segment. It plays a positive and negative two-way adjustment role in correcting stability.

[0063] In practical implementation, when determining the quality stability of the target pesticide formulation during the preparation process based on the corrected stability, the corrected stability of all candidate response offset segments is obtained. The corrected stability is obtained by multiplying the weighted stability of the candidate response offset segments by a trend correction factor. The preset corrected stability threshold is set to 0.35. The basis for setting the corrected stability threshold to 0.35 is that the corrected stability is in the range of 0 to 1. When the corrected stability is lower than 0.35, the candidate response offset segment is still at a low stability level after considering the correction effects of upstream and downstream related segments, and is judged as an abnormal state, which can accurately identify the source of quality offset.

[0064] The corrected stability of each candidate response offset segment is compared with a preset corrected stability threshold. The comparison is performed segment by segment, comparing the corrected stability of each segment with the corrected stability threshold. When the corrected stability of all candidate response offset segments is greater than or equal to the corrected stability threshold, the quality stability of the target pesticide formulation during preparation is determined to be normal.

[0065] When the corrected stability of at least one candidate response offset segment is less than the corrected stability threshold, the quality stability of the target pesticide formulation during preparation is determined to be abnormal. Candidate response offset segments with corrected stability less than the corrected stability threshold are marked as quality offset tracing segments. After marking the quality offset tracing segments, the sensing and monitoring system records the corresponding monitoring segment number, corrected stability value, and determination time into the system's storage unit.

[0066] In practical implementation, the quality stability monitoring sensing system for pesticide preparation includes a memory, a processor, and a computer program stored in the memory and running on the processor. The memory stores the computer program, acoustic and vibration response signals, medium density response signals, disturbance propagation time-series characteristics, density distribution gradient characteristics, segment response stability, candidate response offset segment marking results, corrected stability, and quality stability status determination results collected by the sensing monitoring system. The processor executes the instructions in the computer program. The memory and processor are connected via a data bus, and multiple sets of sensor acquisition nodes in the sensing monitoring system are connected to the processor via data interfaces.

[0067] When executing a computer program, the processor implements the steps of a quality stability monitoring method for pesticide preparation. The implementation steps include: during the preparation of the target pesticide formulation, acquiring acoustic and vibration response signals and medium density response signals of corresponding monitoring sections in the preparation pipeline through various sensing nodes; extracting the disturbance propagation time-series characteristics of each monitoring section based on the acoustic and vibration response signals, and extracting the density distribution gradient characteristics of each monitoring section based on the medium density response signals; determining the segment response stability of each monitoring section based on the disturbance propagation time-series characteristics and density distribution gradient characteristics; marking candidate response offset sections in each monitoring section based on the segment response stability; performing a weighted correction on the segment response stability of the candidate response offset sections based on the upstream and downstream correlation of the candidate response offset sections in the preparation pipeline, obtaining the corrected stability of each candidate response offset section; and determining the quality stability status of the target pesticide formulation during the preparation process based on the corrected stability.

[0068] When the processor executes the computer program to determine the quality stability status of the target pesticide formulation during the preparation process based on the modified stability, it calculates the quality stability status indicator variable according to the following determination formula:

[0069]

[0070] in, This indicates the quality stability state indicator variable, when A value of 1 indicates that the quality stability of the target pesticide formulation is abnormal during the preparation process. When the value is 0, it indicates that the quality stability of the target pesticide formulation is in a normal state during the preparation process; Indicates the first Corrected stability of each candidate response offset segment The sequential numbering of the candidate response offset segment when it is marked; This indicates the corrected stability threshold, which is set to 0.35. This indicates that there exists at least one candidate response offset segment that satisfies the condition that the corrected stability is less than the corrected stability threshold; This indicates that all candidate response offset segments satisfy the condition that the corrected stability is greater than or equal to the corrected stability threshold.

[0071] When the determination formula is calculated When the value is 1, the processor marks candidate response offset segments with a corrected stability value less than the corrected stability threshold as mass offset tracing segments. The processor records the monitoring segment number, corrected stability value, and decision time corresponding to the mass offset tracing segment into memory. When the decision formula calculates... When the value is 0, the processor records the quality stability status of the target pesticide formulation during the preparation process as normal in the memory.

[0072] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for monitoring the quality stability of pesticide preparation, characterized in that, The method is applied to a sensing and monitoring system in a pesticide production line. The sensing and monitoring system includes multiple sets of sensing and data acquisition nodes arranged sequentially along the production pipeline. The method includes: During the preparation of the target pesticide formulation, acoustic vibration response signals and medium density response signals of the corresponding monitoring sections in the preparation pipeline are collected through each of the aforementioned sensing acquisition nodes; The disturbance propagation time sequence characteristics of each monitoring section are extracted based on the acoustic and vibration response signals of each monitoring section, and the density distribution gradient characteristics of each monitoring section are extracted based on the medium density response signals of each monitoring section. Based on the disturbance propagation time sequence characteristics and density distribution gradient characteristics of each monitoring segment, the segment response stability of each monitoring segment is determined; Based on the segment response stability of each monitoring segment, candidate response offset segments are marked in each monitoring segment; Based on the upstream and downstream correlation of the candidate response offset segments in the preparation pipeline, the segment response stability of the candidate response offset segments is weighted and corrected to obtain the corrected stability of each candidate response offset segment. The quality stability of the target pesticide formulation during the preparation process is determined based on the modified stability.

2. The method for monitoring the quality stability of pesticide preparation according to claim 1, characterized in that, The acoustic vibration response signal and medium density response signal of the corresponding monitoring section in the preparation pipeline are collected through each of the aforementioned sensing acquisition nodes, including: The acoustic vibration sensors in each of the aforementioned sensing acquisition nodes acquire the original acoustic vibration signal of the corresponding monitoring section at a first sampling frequency, and perform frequency band separation processing on the original acoustic vibration signal to obtain the high-frequency acoustic vibration component and the low-frequency acoustic vibration component. The original density signal of the corresponding monitoring section is acquired by the density sensing unit in each of the aforementioned sensing acquisition nodes at a second sampling frequency, wherein the second sampling frequency is lower than the first sampling frequency. The high-frequency acoustic vibration component, the low-frequency acoustic vibration component, and the original density signal are aligned according to the acquisition time and then combined to form the acoustic vibration response signal and the medium density response signal of the corresponding monitoring section.

3. The method for monitoring the quality stability of pesticide preparation according to claim 2, characterized in that, The disturbance propagation time-series characteristics of each monitoring section are extracted based on the acoustic and vibration response signals of each monitoring section, including: For any monitoring segment, extract the peak amplitude sequence of the high-frequency acoustic vibration component within multiple acquisition windows from the acoustic vibration response signal of the monitoring segment, and extract the zero-crossing rate sequence of the low-frequency acoustic vibration component within the multiple acquisition windows. Calculate the cross-correlation offset between the peak amplitude sequence and the zero-crossing rate sequence on the time axis; The disturbance propagation delay of the monitoring section is determined based on the cross-correlation offset, and the disturbance propagation delay is used as the disturbance propagation timing feature of the monitoring section.

4. The method for monitoring the quality stability of pesticide preparation according to claim 2, characterized in that, Extracting the density distribution gradient characteristics of each monitoring section based on the medium density response signal of each monitoring section includes: For any monitoring section, the medium density response signal of the monitoring section is divided into multiple density sampling sequences according to the acquisition time, and each density sampling sequence corresponds to an acquisition period; For any density sampling sequence, calculate the density change slope between adjacent sampling points in the density sampling sequence, and arrange all the calculated density change slopes in chronological order to obtain the slope change trajectory of the sampling period. Calculate the slope span between the maximum and minimum slope values ​​in the slope change trajectory, and use the slope span value of this acquisition period as the density distribution gradient feature of the monitoring segment in this acquisition period.

5. The method for monitoring the quality stability of pesticide preparation according to claim 1, characterized in that, Based on the disturbance propagation time series characteristics and density distribution gradient characteristics of each monitoring segment, the segment response stability of each monitoring segment is determined, including: For any given monitoring segment, the disturbance delay fluctuation amplitude of that monitoring segment within multiple consecutive acquisition cycles is determined based on the disturbance propagation timing characteristics of that monitoring segment. Based on the density distribution gradient characteristics of the monitoring section, determine the magnitude of density gradient change in the monitoring section over the consecutive multiple acquisition cycles. The disturbance delay fluctuation amplitude and the density gradient change amplitude are fused according to a preset weight to obtain the segment response stability of the monitoring segment in the continuous multiple acquisition cycles. The segment response stability is negatively correlated with the disturbance delay fluctuation amplitude and negatively correlated with the density gradient change amplitude.

6. The method for monitoring the quality stability of pesticide preparation according to claim 1, characterized in that, Based on the segment response stability of each monitoring segment, candidate response offset segments are marked in each monitoring segment, including: According to the arrangement order of each monitoring section in the preparation pipeline, the section response stability of each monitoring section is arranged into a stability sequence; A local minimum point is determined in the stability sequence, wherein the segment response stability of the local minimum point is less than the response stability of its two adjacent segments in the stability sequence; When the segment response stability of the local minimum point is lower than the preset lower stability threshold, the monitoring segment corresponding to the local minimum point is marked as a candidate response offset segment.

7. The method for monitoring the quality stability of pesticide preparation according to claim 1, characterized in that, Based on the upstream and downstream correlation of the candidate response offset segments in the fabrication pipeline, the segment response stability of the candidate response offset segments is weighted and corrected to obtain the corrected stability of each candidate response offset segment, including: For any candidate response offset segment, determine the upstream and downstream adjacent monitoring segments of that candidate response offset segment; Obtain the first segment response stability of the upstream adjacent monitoring segment and the second segment response stability of the downstream adjacent monitoring segment; Based on the degree of difference between the response stability of the first segment and the response stability of the second segment, a target associated segment with a related fluctuation trend that is associated with the candidate response offset segment is determined among the upstream adjacent monitoring segment and the downstream adjacent monitoring segment; The segment response stability of the candidate response offset segment is weighted and corrected based on the segment response stability of the target associated segment to obtain the corrected stability of the candidate response offset segment.

8. The method for monitoring the quality stability of pesticide preparation according to claim 7, characterized in that, Based on the degree of difference between the response stability of the first segment and the response stability of the second segment, the target associated segments among the upstream adjacent monitoring segments and the downstream adjacent monitoring segments that have a correlated fluctuation trend with the candidate response offset segment are determined as follows: Calculate the absolute difference between the response stability of the first segment and the response stability of the second segment; When the absolute difference is less than a preset difference threshold, both the upstream adjacent monitoring segment and the downstream adjacent monitoring segment are taken as target associated segments. When the absolute difference is greater than or equal to the preset difference threshold, the monitoring segment corresponding to the segment response stability of the candidate response offset segment that has the smaller absolute difference between the first segment response stability and the second segment response stability is taken as the target associated segment. The method for obtaining the corrected stability of the candidate response offset segment specifically includes: Obtain the pipeline distance between the candidate response offset segment and its target associated segment, and determine the stability weight coefficient of the target associated segment based on the pipeline distance. The stability weight coefficient is negatively correlated with the pipeline distance. The segment response stability of the candidate response offset segment is weighted and summed with the segment response stability of the target associated segment according to the stability weighting coefficient to obtain the weighted stability of the candidate response offset segment. Based on the changing trend of the segment response stability of multiple monitoring segments in the upstream direction of the candidate response offset segment, the trend correction factor of the candidate response offset segment is determined. Multiplying the weighted stability by the trend correction factor yields the corrected stability of the candidate response offset segment.

9. The method for monitoring the quality stability of pesticide preparation according to claim 1, characterized in that, Determining the quality stability status of the target pesticide formulation during the preparation process based on the modified stability includes: Compare the corrected stability of each candidate response offset segment with a preset corrected stability threshold; When the corrected stability of all candidate response offset segments is greater than or equal to the corrected stability threshold, the quality stability of the target pesticide formulation during the preparation process is determined to be normal. When the corrected stability of at least one candidate response offset segment is less than the corrected stability threshold, the quality stability state of the target pesticide formulation during the preparation process is determined to be abnormal, and the candidate response offset segment with the corrected stability less than the corrected stability threshold is marked as a quality offset tracing segment.

10. A quality stability monitoring sensing system for pesticide preparation, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the quality stability monitoring method for pesticide preparation as described in any one of claims 1 to 9.

Images