Near-infrared spectroscopy-based analysis system for cotton blending uniformity in multi-compartment cotton cleaning machines

Abstract

This invention relates to the field of near-infrared analysis technology, specifically to a system for analyzing the uniformity of multi-bin cotton blending in a cotton cleaning machine based on near-infrared spectroscopy. The system includes: acquiring near-infrared spectral curves and cotton bin humidity at different sampling times; analyzing the effects of baseline shifting and curve scaling based on fluctuations in the spectral curves to identify abnormal curves; determining the scattering coefficients of abnormal states based on the abnormal state parameters and humidity changes of all abnormal curves; removing the wavelength range absorbed by hydroxyl bonds from the abnormal curves to obtain de-hydroxyl bond curves; determining the degree of influence of hydroxyl bond absorption effect by combining the difference in scattering coefficients between the abnormal curves and the corresponding de-hydroxyl bond curves; and finally, adjusting the ventilation efficiency in the cotton bins according to the degree of influence to achieve uniformity analysis. This invention can effectively analyze the impact of cotton bin humidity on noise interference generated during spectral detection, improving the accuracy of detection results when detecting the uniformity of multi-bin cotton blending.

Description

Technical Field

[0001] This invention relates to the field of near-infrared analysis technology, specifically to a multi-compartment cotton blending uniformity analysis system based on near-infrared spectroscopy for cotton cleaning machines. Background Technology

[0002] The cotton blending process is a crucial intermediate step in the opening and cleaning process. The quality of the blending directly impacts the quality of the sliver and the yarn. Poor blending makes it difficult to fully mix the smaller fiber bundles after opening and impurity removal in subsequent processes. When using a multi-compartment cotton blending system in a cotton cleaning machine to blend the collected cotton, a certain level of moisture is required after washing, spraying, and drying to prevent breakage during the blending process. This ensures the cotton's resilience and maintains a certain level of humidity within the cotton bins. Simultaneously, a fan delivers airflow to the perforated plates between the cotton bins, ensuring that some moisture is carried away as the cotton accumulates, preventing the humidity within the bins from affecting the accuracy of near-infrared spectroscopy data for cotton flow detection.

[0003] When using near-infrared spectroscopy to detect the effect of cotton blending, abnormal wind speed at the perforated plate, such as excessively slow wind speed causing gas stagnation and increasing humidity in the cotton bin, can alter the near-infrared spectral characteristic absorption peaks of the fibers (such as OH bonds). This can cause the detection algorithm to misidentify the noise interference as impurities, resulting in low accuracy of the detection results when using near-infrared spectroscopy to detect the uniformity of cotton blending in multiple bins. Summary of the Invention

[0004] To address the technical problem of low accuracy in multi-bin cotton blend uniformity detection using near-infrared spectroscopy algorithms due to noise generated during cotton bin humidity spectral detection, this invention provides a multi-bin cotton blend uniformity analysis system based on near-infrared spectroscopy for cotton cleaning machines. The specific technical solution adopted is as follows:

[0005] This invention proposes a near-infrared spectroscopy-based system for analyzing the uniformity of cotton blending in multi-compartment cotton cleaning machines. The system includes:

[0006] The acquisition module is used to detect the cotton blend in the cotton warehouse using near-infrared spectroscopy at different sampling times, acquire the near-infrared spectral curve, with the horizontal axis representing wavelength and the vertical axis representing the detected absorbance; determine the humidity of the cotton warehouse at each sampling time; and integrate all spectral curves to obtain the baseline curve.

[0007] The filtering module is used to determine the characteristic peak difference index that causes baseline shift in the spectral curve based on the difference between the spectral curve and the reference curve; to determine the abnormal state parameters of the spectral curve based on the characteristic peak difference index and the absorbance fluctuation changes between the spectral curve and the reference curve; and to filter abnormal curves from all spectral curves based on the abnormal state parameters.

[0008] The absorption effect module is used to determine the scattering coefficient of the scattering anomaly state based on the abnormal state parameters and humidity changes of all abnormal curves; remove the wavelength range absorbed by the hydroxyl bond in the abnormal curve to obtain the de-hydroxyl bond curve; and determine the degree of influence of the hydroxyl bond absorption effect by combining the difference in scattering coefficient between the abnormal curve and the corresponding de-hydroxyl bond curve.

[0009] The analysis module is used to adjust the ventilation efficiency in the cotton silo according to the degree of impact, thereby achieving uniformity analysis.

[0010] Furthermore, the process of integrating all spectral curves to obtain a reference curve includes:

[0011] Calculate the average absorbance at the same wavelength in all spectral curves as the baseline absorbance;

[0012] Connect the reference absorbance coordinates of all wavelengths sequentially in the coordinate system corresponding to the spectral curve to obtain the reference curve.

[0013] Furthermore, determining the characteristic peak difference index that causes baseline shift in the spectral curve based on the difference between the spectral curve and the reference curve includes:

[0014] Calculate the absolute value of the difference between the detected absorbance and the corresponding reference absorbance at the same wavelength, and normalize the mean of all the absolute values ​​of the difference to obtain the characteristic peak difference index.

[0015] Furthermore, determining the abnormal state parameters of the spectral curve based on the characteristic peak difference index and the absorbance fluctuation changes between the spectral curve and the reference curve includes:

[0016] Based on the difference in absorbance fluctuations at all wavelengths between the spectral curve and the reference curve, determine the degree of scaling of the spectral curve.

[0017] The product of the scaling degree and the characteristic peak difference index is normalized and used as the abnormal state parameter of the spectral curve.

[0018] Furthermore, determining the degree of scaling of the spectral curve based on the absorbance fluctuation differences between the spectral curve and the reference curve at all wavelengths includes:

[0019] The standard deviation of the detected absorbance at all wavelengths in the spectral curve is calculated as the first standard deviation;

[0020] The standard deviation of the reference absorbance at all wavelengths in the reference curve is calculated as the second standard deviation;

[0021] The absolute value of the difference between the first and second standard deviations is normalized and used as the scaling factor.

[0022] Furthermore, the step of filtering abnormal curves from all spectral curves based on abnormal state parameters includes:

[0023] The spectral curves whose abnormal state parameters are greater than a preset parameter threshold are considered as abnormal curves.

[0024] Furthermore, determining the scattering coefficient indicating an abnormal scattering state based on the abnormal state parameters and humidity changes of all abnormal curves includes:

[0025] A two-dimensional coordinate system is constructed with the humidity of the abnormal curve as the abscissa and the abnormal state parameters of the abnormal curve as the ordinate, and the coordinate points of the analytical parameters under the humidity of the abnormal curve are determined.

[0026] Perform power function curve fitting on all coordinate points to obtain the fitted curve;

[0027] Calculate the mean distance from each coordinate point on the same horizontal axis to the fitted curve, and normalize the mean distance as the scattering coefficient.

[0028] Further, the removal of the wavelength range absorbed by the hydroxyl bonds in the abnormal curve to obtain the dehydroxyl bond curve includes:

[0029] Calculate the average absorbance of the wavelength range absorbed by the hydroxyl bond, and use it as the average absorbance of hydroxyl.

[0030] The absorbance of the wavelength range absorbed by the hydroxyl bond is adjusted to the average absorbance of hydroxyl bonds to obtain the dehydrogenation curve.

[0031] Furthermore, the determination of the extent of the influence of the hydroxyl bond absorption effect by combining the differences in scattering coefficients between all anomalous curves and all dehydrogenation curves includes:

[0032] The difference between the scattering coefficient of each anomalous curve and its corresponding dehydrogenation curve is taken as the hydroxyl bond influence coefficient of the anomalous curve.

[0033] The mean of the hydroxyl bond influence coefficients of all abnormal curves is normalized to represent the degree of influence.

[0034] Furthermore, a power-adjustable fan is installed at the perforated plate of the cotton silo. Adjusting the ventilation efficiency within the cotton silo according to the degree of impact includes:

[0035] Calculate the product of the impact level and the rated power of the wind turbine as the target power, and adjust the current power of the wind turbine to the target power.

[0036] The present invention has the following beneficial effects:

[0037] In this embodiment of the invention, to address the issue of varying humidity sensitivity, near-infrared spectroscopy is performed on the blended cotton to obtain its spectral curve. Then, by combining the difference between the spectral curve and the reference curve, the characteristic peak difference index causing baseline shift in the spectral curve is determined. Furthermore, signal scaling analysis is performed by combining the absorbance fluctuations of the spectral curve and the reference curve. These two analyses are used to screen out abnormal curves, which represent curves with significant humidity-related spectral differences. Screening out abnormal curves for subsequent analysis effectively amplifies the abnormal effect and improves detection accuracy. Subsequently, the dehydrogenation bond absorption wavelength range is used to obtain the dehydrogenation bond curve analysis. By combining the difference in scattering coefficients between the abnormal curve and the corresponding dehydrogenation bond curve, the degree of influence of the dehydrogenation bond absorption effect is determined, thus accurately determining the impact of humidity on the spectral curve. A greater influence leads to lower accuracy in the obtained uniformity detection results. Therefore, by adjusting the ventilation efficiency in the cotton silo according to the degree of influence, uniformity analysis is achieved, thereby stabilizing and reducing the influence of humidity, resulting in more accurate uniformity analysis results. In summary, the embodiments of the present invention can effectively analyze the impact of cotton warehouse humidity on noise interference generated in spectral detection, and improve the accuracy of detection results when using near-infrared spectroscopy algorithm to detect the uniformity of multi-warehouse cotton blends. Attached Figure Description

[0038] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0039] Figure 1 This is a structural diagram of a multi-compartment cotton blending uniformity analysis system based on near-infrared spectroscopy provided in an embodiment of the present invention. Detailed Implementation

[0040] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a near-infrared spectroscopy-based multi-compartment cotton blending uniformity analysis system for cotton cleaning machines proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0042] The following description, in conjunction with the accompanying drawings, details the specific scheme of the multi-compartment cotton blending uniformity analysis system for a cotton cleaning machine based on near-infrared spectroscopy provided by this invention.

[0043] Please see Figure 1 The diagram illustrates a structural diagram of a multi-compartment cotton blending uniformity analysis system based on near-infrared spectroscopy provided in an embodiment of the present invention. The system includes:

[0044] The acquisition module 101 is used to detect the cotton blend in the cotton warehouse using near-infrared spectroscopy at different sampling times, acquire the spectral curve of the near-infrared spectrum, with the horizontal axis of the spectral curve representing wavelength and the vertical axis representing the detection absorbance; determine the humidity of the cotton warehouse at each sampling time; and integrate all spectral curves to obtain the baseline curve.

[0045] The cotton blending process is a crucial intermediate step in the opening and cleaning process. The quality of the blending directly impacts the quality of the sliver and the yarn. Poor blending makes it difficult to fully mix the smaller fiber bundles after opening and impurity removal in subsequent processes. When using a multi-compartment cotton blending system in a cotton cleaning machine to blend the collected cotton, a certain level of moisture is required after washing, spraying, and drying to prevent breakage during the blending process. This ensures the cotton's resilience and maintains a certain level of humidity within the cotton bins. Simultaneously, a fan delivers airflow to the perforated plates between the cotton bins, ensuring that some moisture is carried away as the cotton accumulates, preventing the humidity within the bins from affecting the accuracy of near-infrared spectroscopy data for cotton flow detection.

[0046] When using near-infrared spectroscopy to detect the effect of cotton blending, abnormal wind speed at the perforated plate, such as excessively slow wind speed causing gas stagnation and increasing humidity in the cotton bin, can alter the near-infrared spectral characteristic absorption peaks of the fibers (such as OH bonds). This can cause the detection algorithm to misidentify the noise interference as impurities, resulting in low accuracy of the detection results when using near-infrared spectroscopy to detect the uniformity of cotton blending in multiple bins.

[0047] When performing near-infrared spectroscopy on multi-compartment blended cotton, the chemical components in the cotton not only absorb near-infrared light but also produce strong scattering, leading to shifts and scaling of the acquired spectral signals. This masks the true chemical absorption signals, resulting in low accuracy of the detection results. This invention addresses these problems by analyzing the temporal changes in near-infrared spectra and humidity.

[0048] This invention relates to a method for installing sensors inside a multi-compartment cotton blending device in a cotton cleaning machine and uploading monitoring data to a uniformity analysis system via a wireless data transmission module. The uploaded monitoring data is pre-processed to remove noise interference.

[0049] For the cotton blend processed by the stripping roller, before conveying the blend to the next device through the feeding pipeline, near-infrared spectroscopy is used to detect the blend at different sampling times to obtain near-infrared spectral curves. A humidity sensor is used to obtain the humidity data of the cotton bin at the sampling times.

[0050] During the monitoring process, the near-infrared spectrum scatters differently at different wavelengths. Therefore, it is necessary to integrate different spectral curves to form a standard reference curve. In this embodiment of the invention, the integration is mean fitting.

[0051] Furthermore, in some embodiments of the present invention, all spectral curves are integrated to obtain a reference curve, including: calculating the average absorbance at the same wavelength in all spectral curves as the reference absorbance; and sequentially connecting the reference absorbance coordinate points of all wavelengths in the coordinate system corresponding to the spectral curves to obtain the reference curve. The reference curve obtained by mean fitting can effectively eliminate the influence of differences in scattering at different wavelengths.

[0052] The screening module 102 is used to determine the characteristic peak difference index that causes baseline shift in the spectral curve based on the difference between the spectral curve and the reference curve; to determine the abnormal state parameter of the spectral curve based on the characteristic peak difference index and the absorbance fluctuation change between the spectral curve and the reference curve; and to screen abnormal curves from all spectral curves based on the abnormal state parameter.

[0053] Near-infrared spectroscopy technology uses infrared spectra of different wavelengths to detect the blend of cotton in a cotton silo. It analyzes the absorption state of near-infrared light by the cotton stream, thereby detecting the content of the main substances (components) in the cotton stream through transmitted light. This allows for the detection of the component content of different types of cotton within the silo, thus enabling the assessment of the uniformity of the cotton blend. However, during the detection process, cotton not only absorbs near-infrared light but also undergoes a physical reaction—scattering. This scattering is easily affected by humidity, resulting in spectral baseline shifts and scaling in the spectral data obtained from detecting the cotton stream.

[0054] Baseline shifting causes the entire spectral curve to move up and down along the vertical axis (absorbance), while scaling compresses or stretches the amplitude of the spectral curve, thus masking the true absorption information and reducing the model's prediction accuracy and robustness. Therefore, it is necessary to screen the raw data to identify anomalous spectral curve segments affected by scattering in the spectral detection.

[0055] The greater the difference between the spectral curve and the reference curve, the more severe the baseline shift. The higher the degree of baseline shift, the more severe the baseline shift. Therefore, this embodiment of the invention uses the characteristic peak difference index to characterize the degree of baseline shift.

[0056] Furthermore, in some embodiments of the present invention, the characteristic peak difference index of the spectral curve that produces baseline shift is determined based on the difference between the spectral curve and the reference curve, including: calculating the absolute value of the difference between the detected absorbance and the corresponding reference absorbance at the same wavelength, and normalizing the mean of all the absolute values ​​of the difference to obtain the characteristic peak difference index.

[0057] Among them, the larger the absolute value of the difference between the detected absorbance and the corresponding reference absorbance at the same wavelength, the greater the difference between the detected absorbance and the reference absorbance. Then, all the absolute values ​​of the difference are integrated, and the mean of all the absolute values ​​of the difference is normalized to obtain the characteristic peak difference index.

[0058] In one embodiment of the present invention, the normalization process can be specifically, for example, maximum and minimum value normalization. Furthermore, the normalization in subsequent steps can all adopt maximum and minimum value normalization. In other embodiments of the present invention, other normalization methods can be selected according to the specific range of the numerical values, which will not be elaborated further.

[0059] In near-infrared spectroscopy analysis, anomalies caused by physical scattering of cotton not only manifest as baseline shifts but are also often accompanied by signal amplitude scaling. This scaling can be analyzed by absorbance fluctuations. Therefore, this embodiment of the invention combines characteristic peak difference indices with absorbance fluctuations between the spectral curve and the reference curve to achieve anomaly analysis. Specifically, the anomaly is the baseline shift and signal scaling caused by scattering.

[0060] Furthermore, in some embodiments of the present invention, the abnormal state parameters of the spectral curve are determined based on the characteristic peak difference index and the absorbance fluctuation changes between the spectral curve and the reference curve, including: determining the scaling degree of the spectral curve affected by scaling based on the absorbance fluctuation differences at all wavelengths of the spectral curve and the reference curve; and normalizing the product of the scaling degree and the characteristic peak difference index as the abnormal state parameters of the spectral curve.

[0061] First, the scaling degree is analyzed. In this embodiment of the invention, the fluctuation difference is specifically the difference in discreteness. By analyzing the fluctuation change of absorbance, the scaling degree of the spectral curve affected by scaling is determined, including: calculating the numerical standard deviation of the detected absorbance of all wavelengths in the spectral curve as the first standard deviation; calculating the numerical standard deviation of the reference absorbance of all wavelengths in the reference curve as the second standard deviation; and normalizing the absolute value of the difference between the first standard deviation and the second standard deviation as the scaling degree.

[0062] In other words, the absolute value of the difference between the numerical standard deviation of the detected absorbance at all wavelengths in the spectral curve and the numerical standard deviation of the reference absorbance at all wavelengths in the reference curve is calculated, and then normalized to obtain the scaling degree. The scaling degree represents the difference in dispersion between the spectral curve and the reference curve. The larger the value, the greater the effect of scaling on the spectral curve.

[0063] The scaling factor indicates the degree to which the spectral curve is affected by scaling, while the characteristic peak difference index indicates the degree of baseline shift. Combining these two parameters enables anomaly analysis. The product of the scaling factor and the characteristic peak difference index is normalized and used as the anomaly parameter of the spectral curve. The anomaly parameter represents the overall analytical result affected by scattering.

[0064] Furthermore, in some embodiments of the present invention, screening abnormal curves from all spectral curves based on abnormal state parameters includes: selecting spectral curves whose abnormal state parameters are greater than a preset parameter threshold as abnormal curves.

[0065] The preset parameter threshold is a threshold value for the abnormal state parameter. In this embodiment of the invention, the preset parameter threshold can be, for example, 0.5, so that the spectral curve with the abnormal state parameter greater than 0.5 is regarded as the abnormal curve. In other embodiments of the invention, the preset parameter threshold can also be adjusted according to the actual detection scenario, and there is no limitation on this.

[0066] The absorption effect module 103 is used to determine the scattering coefficient of the scattering anomaly state based on the abnormal state parameters and humidity changes of all abnormal curves; remove the wavelength range absorbed by the hydrogen-oxygen bond in the abnormal curve to obtain the dehydrogen-oxygen bond curve; and determine the degree of influence of the hydrogen-oxygen bond absorption effect by combining the difference in scattering coefficient between the abnormal curve and the corresponding dehydrogen-oxygen bond curve.

[0067] Cotton continuously enters the cotton bin through the inlet pipe. When the cotton contains large impurities or a large amount of cotton rushes in at a single moment, it can clog the channels, affecting the airflow and preventing excess moisture from escaping. This increases the moisture content of the cotton flow within the bin. Increased humidity causes moisture to adhere to the cotton surface, increasing light scattering between particles. This can lead to baseline shifting and magnification issues when using near-infrared light technology to detect the cotton flow, affecting the actual monitoring accuracy.

[0068] Based on the abnormal state parameters and humidity changes of all abnormal curves, the scattering coefficients of the abnormal scattering states are determined, including: constructing a two-dimensional coordinate system with the humidity of the abnormal curve as the abscissa and the abnormal state parameters of the abnormal curve as the ordinate, and determining the coordinate points of the analysis parameters under the humidity of the abnormal curve; performing power function curve fitting on all coordinate points to obtain the fitted curve; calculating the mean distance from each coordinate point to the fitted curve under the same abscissa, and normalizing the mean distance as the scattering coefficient.

[0069] Due to the differences in the adhesion and absorption of moisture by cotton under different humidity conditions, the baseline shift and scaling caused by the scattering of near-infrared light are non-linear changes. In this embodiment of the invention, a power function is used as the curve fitting result.

[0070] The scattering coefficient represents the difference between the scattering situation under increased humidity and the standard situation. The larger the value of the scattering coefficient, the greater the abnormality of the cotton blend in the corresponding cotton warehouse affected by scattering. This abnormality will lead to a greater deviation in the uniformity analysis results and a lower prediction accuracy.

[0071] The high humidity within the cotton silo results in a high content of hydroxyl (OH) bonds, leading to excessively high absorption of the cotton at the wavelengths corresponding to these bonds during near-infrared light detection. Since the increased humidity in the silo primarily affects the hydroxyl bonds, causing significant variations in the absorbance of the cotton at these wavelengths, the influence of hydroxyl bonds is eliminated to analyze the differences in the actual spectral curves.

[0072] Furthermore, in some embodiments of the present invention, removing the wavelength range absorbed by hydroxyl bonds in the abnormal curve to obtain a de-hydroxyl bond curve includes: calculating the average value of the detected absorbance of the wavelength range absorbed by hydroxyl bonds as the hydroxyl absorbance average; adjusting the detected absorbance of the wavelength range absorbed by hydroxyl bonds to the hydroxyl absorbance average to obtain the de-hydroxyl bond curve.

[0073] In this embodiment of the invention, the wavelength range absorbed by the hydroxyl bond is a preset wavelength range based on the actual situation. By averaging, the fluctuation characteristics of the wavelength range absorbed by the hydroxyl bond are eliminated while retaining the numerical magnitude characteristics, thereby obtaining the remaining dehydrogenation curve for analysis.

[0074] In other words, in this embodiment of the invention, the dehydrogenation curve is the near-infrared spectral information obtained after eliminating the influence of humidity. Hydrogenation bond influence analysis is performed by using the dehydrogenation curve and the abnormal curve, and the scattering coefficient of the dehydrogenation curve of all abnormal curves is calculated by analogy.

[0075] Furthermore, in some embodiments of the present invention, the degree of influence of the hydroxyl bond absorption effect is determined by combining the differences in scattering coefficients between all anomalous curves and all dehydrogenation curves, including: taking the difference in scattering coefficients between each anomalous curve and its corresponding dehydrogenation curve as the hydroxyl bond influence coefficient of the anomalous curve; and normalizing the mean of the hydroxyl bond influence coefficients of all anomalous curves as the degree of influence.

[0076] In this embodiment of the invention, the degree of influence is a parameter representing the extent to which hydroxyl bonds affect the system. Since humidity is primarily manifested in the form of hydroxyl bonds, the difference between the scattering coefficient of each anomalous curve and its corresponding dehydrogenation curve is used as the hydroxyl bond influence coefficient of the anomalous curve. The larger the value of the hydroxyl bond influence coefficient, the more significant the influence of humidity. Therefore, the mean of the hydroxyl bond influence coefficients of all anomalous curves is normalized to represent the degree of influence.

[0077] This solution is a multi-compartment cotton blending uniformity analysis system based on near-infrared spectroscopy for cotton cleaning machines. It uses near-infrared spectroscopy to detect the cotton stream after blending, and obtains the absorption effect of different substances in the cotton stream on different wavelengths after being irradiated with near-infrared light. Based on the absorbance of characteristic peaks in the spectrum, the components of the cotton stream are detected. However, due to the different sensitivities of different cotton materials to humidity, the scattering of near-infrared light by the cotton materials and the difference in the degree of scattering caused by humidity changes result in low accuracy of the actual detection results.

[0078] Therefore, the degree of influence calculated in the embodiments of the present invention represents the adaptive influence of humidity on near-infrared light under the detected cotton material, thereby obtaining a more matching and accurate influence analysis, which facilitates subsequent ventilation analysis based on the degree of influence.

[0079] Analysis module 104 is used to adjust the ventilation efficiency in the cotton silo according to the degree of influence, so as to achieve uniformity analysis.

[0080] Furthermore, in some embodiments of the present invention, a power-adjustable fan is provided at the cotton silo perforated plate, the product of the degree of influence and the rated power of the fan is calculated as the target power, and the current power of the fan is adjusted to the target power.

[0081] In other words, the degree of influence is directly used as the linear adjustment weight of the rated power of the fan. Of course, in other embodiments of the present invention, under the adjustment logic that the greater the degree of influence, the greater the rated power of the fan, the fan power can be adjusted according to the actual situation, and the fan power will gradually increase with the increase of humidity to achieve more accurate and reliable control, thereby improving ventilation efficiency, stabilizing the cotton silo detection environment, and obtaining more accurate and reliable uniformity analysis results.

[0082] In this embodiment of the invention, to address the issue of varying humidity sensitivity, near-infrared spectroscopy is performed on the blended cotton to obtain its spectral curve. Then, by combining the difference between the spectral curve and the reference curve, the characteristic peak difference index causing baseline shift in the spectral curve is determined. Furthermore, signal scaling analysis is performed by combining the absorbance fluctuations of the spectral curve and the reference curve. These two analyses are used to screen out abnormal curves, which represent curves with significant humidity-related spectral differences. Screening out abnormal curves for subsequent analysis effectively amplifies the abnormal effect and improves detection accuracy. Subsequently, the dehydrogenation bond absorption wavelength range is used to obtain the dehydrogenation bond curve analysis. By combining the difference in scattering coefficients between the abnormal curve and the corresponding dehydrogenation bond curve, the degree of influence of the dehydrogenation bond absorption effect is determined, thus accurately determining the impact of humidity on the spectral curve. A greater influence leads to lower accuracy in the obtained uniformity detection results. Therefore, by adjusting the ventilation efficiency in the cotton silo according to the degree of influence, uniformity analysis is achieved, thereby stabilizing and reducing the influence of humidity, resulting in more accurate uniformity analysis results. In summary, the embodiments of the present invention can effectively analyze the impact of cotton warehouse humidity on noise interference generated in spectral detection, and improve the accuracy of detection results when using near-infrared spectroscopy algorithm to detect the uniformity of multi-warehouse cotton blends.

[0083] It should be noted that the order of the above embodiments of the present invention is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0084] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A near-infrared spectrum-based analysis system for evenness of multi-bunker mixed cotton in a cotton cleaner, characterized in that, The system comprises: An acquisition module is configured to detect the cotton mixing in the cotton warehouse using near-infrared spectroscopy at different sampling time points, to acquire a spectrum curve of the near-infrared spectroscopy, the abscissa of the spectrum curve being wavelength and the ordinate being detection absorbance, to determine the humidity of the cotton warehouse at each sampling time point, and to integrate all the spectrum curves to obtain a reference curve; A screening module is configured to determine a characteristic peak difference index of baseline translation of the spectrum curve according to the difference between the spectrum curve and the reference curve, to determine an abnormal state parameter of the spectrum curve according to the characteristic peak difference index and the absorbance fluctuation change of the spectrum curve and the reference curve, and to screen an abnormal curve from all the spectrum curves according to the abnormal state parameter; An absorption influence module is configured to determine a scattering coefficient of the scattering abnormal state according to the abnormal state parameter and the humidity change of all the abnormal curves, to remove the wavelength segment absorbed by the hydrogen bond in the abnormal curve to obtain a hydrogen bond-removed curve, and to determine the influence degree of the hydrogen bond absorption effect by combining the scattering coefficient difference between the abnormal curve and the corresponding hydrogen bond-removed curve; An analysis module is configured to adjust the ventilation efficiency in the cotton warehouse according to the influence degree and to realize uniformity analysis; The characteristic peak difference index of baseline translation of the spectrum curve is determined according to the difference between the spectrum curve and the reference curve, and the characteristic peak difference index is obtained by normalizing the mean value of the absolute value of the difference between the detection absorbance and the corresponding reference absorbance at the same wavelength. The abnormal state parameter of the spectrum curve is determined according to the characteristic peak difference index and the absorbance fluctuation change of the spectrum curve and the reference curve, and the abnormal state parameter of the spectrum curve is obtained by normalizing the product value of the scaling degree and the characteristic peak difference index. The scaling degree of the spectrum curve affected by scaling is determined according to the absorbance fluctuation difference of all wavelengths of the spectrum curve and the reference curve, and the scaling degree is obtained by normalizing the absolute value of the difference between the first standard deviation and the second standard deviation. The scattering coefficient of the scattering abnormal state is determined according to the abnormal state parameter and the humidity change of all the abnormal curves, and the scattering coefficient is obtained by normalizing the distance mean value of each coordinate point to the fitting curve at the same abscissa. The wavelength segment absorbed by the hydrogen bond in the abnormal curve is removed to obtain a hydrogen bond-removed curve, and the hydrogen bond-removed curve is obtained by adjusting the detection absorbance of the wavelength segment absorbed by the hydrogen bond to the hydrogen absorption mean value. The scattering coefficient difference between the all abnormal curves and the all dehydrogen-oxygen bond curves determines the influence degree of the hydrogen-oxygen bond absorption effect, comprising: taking the difference between the scattering coefficient of each abnormal curve and the corresponding dehydrogen-oxygen bond curve as the hydrogen-oxygen bond influence coefficient of the abnormal curve; and taking the mean value of the hydrogen-oxygen bond influence coefficients of all abnormal curves as the influence degree after normalization processing.

2. The near infrared spectroscopy based cleaning machine multi-bunker cotton blending uniformity analysis system according to claim 1, characterized in that, The integration of all the spectral curves to obtain the reference curve comprises: Calculating the mean value of the detected absorbance at the same wavelength in all the spectral curves as the reference absorbance; Connecting the reference absorbance coordinate points of all wavelengths in the coordinate system corresponding to the spectral curve in sequence to obtain the reference curve.

3. The near infrared spectroscopy based cleaning machine multi-bunker cotton blending uniformity analysis system according to claim 1, characterized in that, The abnormal curve is selected from all the spectral curves according to the abnormal state parameter, comprising: Taking the spectral curve with the abnormal state parameter greater than the preset parameter threshold as the abnormal curve.

4. The near infrared spectroscopy based cleaning machine multi-bunker cotton blending uniformity analysis system according to claim 1, characterized in that, The cotton bin hole plate is provided with a fan for power adjustment, and the ventilation efficiency in the cotton bin is adjusted according to the influence degree, comprising: Calculating the product value of the influence degree and the rated power of the fan as the target power, and adjusting the current power of the fan to the target power.

Images