A device for measuring the moisture content of an object

Abstract

The application relates to a kind of object mass moisture content measuring devices, including measurement module and analysis module;Measurement module includes cylindrical cavity, capsule, transparent plastic pipeline for capsule passing, two connectors and two probes;Cylindrical cavity includes cavity part and top;Transparent plastic pipeline is along the central axis and passes through cavity part, two connectors are arranged in preset position of top, and one probe is mounted on each connector;Probe is connected with analysis module;Capsule is loaded with the object to be measured;When capsule passes through transparent plastic pipeline, analysis module obtains insertion loss parameter, analyzes insertion loss parameter, and obtains moisture content.It can be understood that the technical scheme shown in the application can non-destructively measure the moisture content of the capsule, and can evaluate the moisture content in the capsule at high speed, which is more advantageous than the traditional method.

Description

Technical Field

[0001] This invention relates to the field of moisture content detection technology, and specifically to a device for measuring the moisture content of an object's mass. Background Technology

[0002] Globally, approximately 200 billion empty capsules are consumed annually, the majority of which are used in pharmaceuticals and health products. Capsule medications account for about 20% of total oral solid dosages, highlighting their important use. Therefore, ensuring the quality control of capsule medications is crucial. Among various quality assessment factors, moisture content and drug concentration within the capsule are key indicators. High moisture content in powdered capsule medications can lead to deliquescence, mold formation, denaturation, or deterioration of drug components, thus affecting their efficacy. Conversely, insufficient moisture content in the drug powder can lead to excessive dryness, thus affecting drug absorption. Furthermore, the drug concentration within the capsule directly impacts the drug's effectiveness and safety.

[0003] Currently, near-infrared reflectance spectroscopy is used to measure drug content, while liquid chromatography is used to separate and quantify the content of five drugs in the study to determine drug content. Additionally, artificial neural networks are used to predict drug content. However, all of the above methods have drawbacks, such as requiring capsule destruction or having insufficient accuracy. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a device for measuring the moisture content of an object, so as to solve the problems of existing methods for measuring moisture content requiring destruction of capsules or insufficient accuracy.

[0005] According to a first aspect of the present invention, an apparatus for measuring the water content of an object's mass is provided, comprising:

[0006] Measurement module and analysis module;

[0007] The measurement module includes a cylindrical cavity, a capsule, a transparent plastic tube for the capsule to pass through, two connectors, and two probes; the cylindrical cavity includes a cavity section and a top; the transparent plastic tube passes through the cavity section along the central axis, and the two connectors are located at preset positions on the top, with a probe mounted on each connector; the probes are connected to the analysis module; the capsule contains the object to be measured.

[0008] When the capsule passes through the transparent plastic tube, the analysis module acquires the insertion loss parameter, analyzes the insertion loss parameter, and obtains the water content.

[0009] Preferably, the connector is symmetrically arranged with respect to the central axis of the cylindrical cavity, and is located on the left and right sides at a quarter position of the central axis.

[0010] Preferably, the cylindrical cavity is made of metal, and the diameter of the cylindrical cavity is 120 mm and the depth is 23 mm.

[0011] Preferably, the analysis module acquires insertion loss parameters, analyzes the insertion loss parameters, and obtains the water content, including:

[0012] The corresponding resonant frequency is obtained based on the insertion loss parameters.

[0013] The dielectric constant corresponding to the resonant frequency is obtained;

[0014] The water content is determined based on the dielectric constant.

[0015] Preferably, when the analysis module derives the corresponding resonant frequency based on the insertion loss parameter, it includes:

[0016] A recognition classifier model is established based on the PCA method and the NB classifier model;

[0017] The insertion loss parameter is identified using the identification classifier model to obtain the corresponding resonant frequency.

[0018] Preferably, when the analysis module derives the water content based on the dielectric constant, it includes:

[0019] The water content can be determined based on the dielectric constant using the following correspondence:

[0020] f r = -0.0173MC + 12.9479

[0021] Among them, f r is the dielectric constant, and MC is the water content.

[0022] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0023] It is understood that the technical solution presented in this invention includes a measurement module and an analysis module. The measurement module includes a cylindrical cavity, a capsule, a transparent plastic tube for the capsule to pass through, two connectors, and two probes. The cylindrical cavity includes a cavity section and a top. The transparent plastic tube passes through the cavity section along its central axis, and the two connectors are positioned at preset locations on the top, with a probe mounted on each connector. The probes are connected to the analysis module. The capsule contains the object to be measured. When the capsule passes through the transparent plastic tube, the analysis module acquires the insertion loss parameter, analyzes the insertion loss parameter, and obtains the water content. It is understood that the technical solution presented in this invention can not only distinguish cases where the relative permittivity difference between the drug and the water inside the capsule is significant, but also effectively distinguish cases where the relative permittivity difference between the drug and the air inside the capsule is small. It can perform non-destructive measurement of the water content of the capsule and can rapidly assess the water content inside the capsule, which is more advantageous than traditional methods.

[0024] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0025] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0026] Figure 1 This is a schematic diagram of an object mass moisture content measuring device according to an exemplary embodiment;

[0027] Figure 2 This is a schematic diagram illustrating a comparison of the number of patterns calculated theoretically and statistically, according to an exemplary embodiment.

[0028] Figure 3 This is a schematic diagram illustrating the relationship between the S21 parameter and the frequency according to an exemplary embodiment;

[0029] Figure 4 This is a schematic diagram illustrating the relationship between frequency and water content according to an exemplary embodiment; Detailed Implementation

[0030] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the invention as detailed in the appended claims.

[0031] Example 1

[0032] Figure 1 This is a schematic diagram of an object mass moisture content measuring device according to an exemplary embodiment, see below. Figure 1 A device for measuring the water content of an object's mass is provided, comprising:

[0033] Measurement module and analysis module;

[0034] The measurement module is mainly a cylindrical cavity testing device, including a cylindrical cavity, a capsule, a transparent plastic tube for the capsule to pass through, two connectors and two probes; the cylindrical cavity includes a cavity section and a top.

[0035] The transparent plastic tube runs through the cavity along the central axis, and the two connectors are located at a preset position at the top, with a probe installed on each connector; the probes are connected to the analysis module; the capsule contains the object to be tested.

[0036] When the capsule passes through the transparent plastic tube, the analysis module acquires the insertion loss parameter, analyzes the insertion loss parameter, and obtains the water content.

[0037] It should be noted that the cylindrical cavity is made of metal, and the diameter of the cylindrical cavity is 120 mm and the depth is 23 mm.

[0038] In practical application, the three-view structure and geometric parameters of the cylindrical cavity are described in [reference needed]. Figure 1 , Figure 1 (a) is a front view. It can be seen that the connectors are symmetrically arranged with respect to the central axis of the cylindrical cavity, and are positioned on either side of the quarter-axis. This arrangement improves the measurement sensitivity for capsules with different moisture or drug contents. Two RF connectors are mounted on the test apparatus through threaded holes with a diameter of φ = 1 / 4 inch. The currently used RF connector is 2.92-KYK, also known as a 2.92 mm coaxial adapter. These adapters offer a wide operating frequency range up to 40 GHz. The RF connectors can accommodate probes similar to monopole antennas, with a diameter of 0.8 mm and an exposed length of 2 mm.

[0039] In addition, a transparent plastic tube with an outer radius of 5 mm, located at the center, passes through the cylindrical cavity device. Throughout the measurement process, capsules with different humidity or drug contents pass through this transparent tube. These capsules have a hollow structure with a shell thickness of 0.1 mm to contain the internal drug powder. Furthermore, the dielectric constant of the drug powder inside the capsule is modulated to simulate different humidity levels. The analysis module can study the relationship between the S21 parameter (insertion loss parameter) and different humidity contents, or classify capsules with different humidity contents using machine learning techniques.

[0040] The technical solution in this embodiment uses the following theoretical basis when performing data analysis:

[0041] The mode analysis and resonant frequency calculation are as follows:

[0042] At lower frequencies, the mode distribution of the electromagnetic field within the cylindrical cavity can lead to inaccuracies in S21 parameter measurements. This inaccuracy can adversely affect the determination of humidity and drug content using the cylindrical cavity method. Therefore, establishing a minimum frequency threshold is crucial for ensuring the reliable implementation of this technique. The total number of modes in the cylindrical cavity depends on the TM solved using the boundary conditions of the circular waveguide. nml and TE nml The resonant frequency of the mode. Equations 1 and 2 respectively give the TM... nml and TE nml The resonant frequency of the mode.

[0043] Formula 1:

[0044] Formula 2:

[0045] Where, p nm and p' nm These correspond to the m-th roots of the Bessel functions Jh(x) and Jh'(x), respectively, where Jh'(x) refers to the derivative of Jh(x). Parameter a represents the radius of the cylindrical cavity, denoted as R4 in this embodiment, while parameter d represents the height of the cylindrical cavity, denoted as H2. Furthermore, μ... r and ε r Let represent the relative permeability and dielectric constant within the cylindrical cavity, respectively, and let c represent the velocity of electromagnetic waves in free space.

[0046] TM 110 The mode exhibits the lowest resonant frequency at 2.3 GHz, indicating that the cylindrical cavity device can operate at frequencies above 2.3 GHz. Based on the definitions of resonant frequency in Equations 1 and 2 mentioned above, the formula for determining the total number of modes in the cylindrical cavity is as follows:

[0047] Equation 3: N≈1.41×10 -29 f 3 -2.88×10 -20 f 2 +4.36×10 -10 f-0.58

[0048] In this test setup, the number of modes N and the frequency f are correlated using Equation 3, which is obtained by fitting data generated using a moment method code. According to Equation 3, the number of modes supported by the test setup increases with increasing frequency. Therefore, the minimum operating frequency in this case is set to 2.3 GHz. The results obtained using Equation 3 are in close agreement with those obtained using Equations 1 and 2, as shown below. Figure 2 As shown, the correlation coefficient of the fitted curve is 0.94.

[0049] The cylindrical cavity exhibits a high electric field intensity region at high frequencies, and energy can be observed to propagate almost uniformly within the test device.

[0050] The skin depth of a material represents the distance a plane wave can penetrate its surface. It can be determined based on the frequency f of the electromagnetic wave, the electrical conductivity σ, and the magnetic permeability μ of the material. The skin depth is calculated as follows:

[0051] Formula 4:

[0052] The electrical conductivity and magnetic permeability of gold are 4.1 × 10⁻⁶. 7 S / m and 1.25659×10 -6 H / m. The cylindrical cavity device operates at frequencies exceeding 10 GHz, and at 10 GHz, the cortical depth does not exceed 7.86 × 10⁻⁶. -7 Meters. Therefore, a metal layer with a thickness greater than 1 μm can be considered an ideal metal.

[0053] Regarding humidity and average resolution, the following is true:

[0054] Humidity (MC) can be expressed as a percentage:

[0055] Formula 5:

[0056] Where, m w The mass of water, m i This represents the mass of each substance. Then, the moisture content of the encapsulated drug can be defined as:

[0057] Formula Six:

[0058] Where, m d This indicates the mass of the drug when it is dried.

[0059] The complex permittivity of the mixture can be accurately calculated using a power-law dielectric mixing model. This model considers the volume weighting of each individual dielectric component, as shown below:

[0060] Formula 7:

[0061] Among them, v m Let β be the volume of the medium, β be a dimensionless parameter, and ε be the volume of the medium. j m j and ρ j These represent the dielectric constant, mass, and density of each medium.

[0062] In this case, the medium mainly consists of the drug, water, and air. Water has a significantly higher relative permittivity than the drug; therefore, as the water content in the drug capsule increases, the overall relative permittivity of the substance also increases. Furthermore, air has a lower relative permittivity than the drug, resulting in a decrease in the overall relative permittivity of the substance when the drug content in the capsule decreases. It can be observed from Equations 1 and 2 that as ε... r The increase of f mnl It will also decrease. This indicates that as the relative permittivity of the medium increases, the resonant frequency will also decrease.

[0063] According to Equations 6 and 7, a correlation can be established between the humidity content of the drug and the dielectric constant of the drug in the capsule, so that the S-parameter simulation process can use machine learning to quantify the difficult-to-express humidity content as an adjustable parameter, the dielectric constant.

[0064] Regarding the theory of principal component analysis:

[0065] Principal Component Analysis (PCA) is an unsupervised feature extraction method that aims to transform multiple correlated variables into a few uncorrelated principal components using dimensionality reduction. These principal components can represent information from the original variables. Notably, each principal component contains unique information, thus reducing feature redundancy. PCA is commonly used for dimensionality reduction of high-dimensional data, thereby simplifying complex tasks. The entire PCA process is as follows:

[0066] Assume the input data is X, an m×n matrix, where m represents the number of input data points and n represents the number of variables in the input data, as shown below:

[0067] Formula 8:

[0068] To eliminate the influence of data dimensionality, the input data needs to be standardized. The standardized data can be calculated as follows:

[0069] Formula Nine:

[0070] in, x represents i average value, Represents x i The variance value.

[0071] Then, the covariance matrix Cov of the standardized data can be calculated using Equation 10:

[0072] Formula 10:

[0073] The covariance matrix is ​​a symmetric matrix. To obtain the projection matrix of the standardized data, it is necessary to calculate the eigenvalues ​​and eigenvectors of the covariance matrix Cov.

[0074] Equation 11: |Cov-λI|=0

[0075] The eigenvalues ​​λ of the covariance matrix Cov can be solved using Equation 11. j And arranged in descending order as λ1≥λ2≥...≥λ n ≥0. The eigenvector corresponding to each eigenvalue can be represented as E=(e1,e2,...,e n Using the example calculation formula twelve for the cumulative contribution rate, principal components can be selected by considering the first few eigenvalues.

[0076] Formula Twelve:

[0077] Normally, M k The value is 80% or higher. Based on the first k principal components, the dimensionality-reduced data can be represented as follows:

[0078] Equation 13: X' = E k X

[0079] Where E k It is (e1,e2,...,e k ).

[0080] In summary, it can be seen that PCA can not only transform raw data into composite variable data that are unrelated to each other, but also minimize the loss of information contained in the raw data, so as to achieve the purpose of comprehensive analysis of the raw data.

[0081] Regarding Naive Bayes theory, the following is stated:

[0082] The Naive Bayes (NB) classifier is a classification algorithm based on Bayes' theorem and has become a mainstream algorithm in machine learning. Due to its stable classification performance, simple process, and significant classification accuracy, the Naive Bayes classification algorithm has been widely used in many fields. In particular, the Naive Bayes classification algorithm requires that the sample features be independent of each other. When calculating the probability of a sample being classified as positive in each class, the class with the highest probability value is selected as the label information for the sample to be classified.

[0083] Given a training set S containing T samples, each sample has q attributes, denoted as {A1, A2, ..., A...} q}, where there are v sample categories, denoted as G={G1,G2,...,G v The number of samples in each category is} j = 1, 2, ..., v. Suppose we have a sample h = (h1, h2, ..., hv) to be judged. q In a given case where the category information is unknown, the Naive Bayes algorithm is used to calculate the conditional probability P(G) that the sample h to be judged belongs to the category. j |h) is as follows:

[0084] Formula Fourteen:

[0085] Since the sample features are independent of each other, Equation 14 can be rewritten as:

[0086] Formula 15:

[0087] in, Represents category G j The prior probability, P(h) represents the prior probability of sample h, P(G) j |h) indicates that it belongs to category G j Under the premise that the sample h to be judged has attribute A i The value above is h i The conditional probability of the sample h belonging to each category G can be calculated based on this. j The probability P j =P(G j |h), j=1,2,...,v, and set the label of the sample h to be judged to have the maximum probability value P. j The category.

[0088] The water content of the capsules is analyzed as follows:

[0089] It should be noted that the analysis module acquires insertion loss parameters, analyzes these parameters, and obtains the water content, including:

[0090] The corresponding resonant frequency is obtained based on the insertion loss parameters.

[0091] The dielectric constant corresponding to the resonant frequency is obtained;

[0092] The water content is determined based on the dielectric constant.

[0093] In practice, high-frequency structural simulation software (HFSS) was used to obtain full-wave simulation results for drug capsules with different water contents. Since differences in water content lead to variations in the relative permittivity, studying different water contents within the capsule can be analogous to analyzing different relative permittivity of the substances within the capsule. Then, the variation in water content was investigated, and the resulting change in the center frequency of the S21 parameter was quantitatively analyzed. Based on these assessments, the sensitivity of the proposed testing method was also evaluated.

[0094] The relationship between dielectric constant and water content can be characterized by equations six and seven. The dielectric constant of a mixture is determined by the dielectric constants of its constituent substances, and the dielectric constant of water is significantly higher than that of the powder in the capsule. Therefore, as the water content increases, the relative dielectric constant of the mixture also increases. The relative dielectric constant is negatively correlated with the resonant frequency. The higher the relative dielectric constant, the lower the resonant frequency. Therefore, as the water content increases, the resonant frequency decreases. Figure 3 The visualization shows the variation of the S21 parameter of the drug in the capsule with frequency, revealing that the relative permittivity varies between 6 and 11. The corresponding center frequencies of the S21 parameter in the drug capsule are as follows: 12.7894 GHz, 12.8134 GHz, 12.8364 GHz, 12.8580 GHz, 12.8800 GHz, and 12.9026 GHz. The difference between the maximum and minimum center frequencies is 113.2 MHz.

[0095] The relative permittivity can be used to determine water content, thereby analyzing the relationship between center frequency and water content. Water content is calculated using the relative permittivity and center frequency values ​​from six different sets of permittivity, such as... Figure 4 As shown, a linear negative correlation was found. The fitting result approximates the following form:

[0096] Formula 16: f r = -0.0173MC + 12.9479

[0097] Therefore, it should be noted that when the analysis module derives the water content based on the dielectric constant, it includes:

[0098] The water content can be determined based on the dielectric constant using the following correspondence:

[0099] f r= -0.0173MC + 12.9479

[0100] Among them, f r is the dielectric constant, and MC is the water content.

[0101] Average resolution (AR) is the average frequency shift caused by fluctuations in the relative permittivity among the M sets of data. This parameter plays a crucial role in evaluating the sensor's sensitivity to changes in the water content of a target substance. In the context of the cylindrical cavity method, the average resolution of the relative permittivity is defined as follows:

[0102] Formula 17:

[0103] in <f k > indicates f k The average of M values, <ε rk > represents ε rk The average of M values ​​is given. Here, M represents the total number of sampling points. It is worth noting that the higher the AR, the greater the sensitivity of the proposed cylindrical cavity fixture.

[0104] According to Formula 17 and Figure 4 The water content and center frequency data shown indicate that the average resolution of the water content relative permittivity in the cylindrical cavity method is 17 MHz percentage MC.

[0105] It should be noted that when the analysis module derives the corresponding resonant frequency based on the insertion loss parameters, it includes:

[0106] A recognition classifier model is established based on the PCA method and the NB classifier model;

[0107] The insertion loss parameter is identified using the identification classifier model to obtain the corresponding resonant frequency.

[0108] In practice, cylindrical cavity fixtures and machine learning can be used to measure and analyze the water content of capsules, as follows:

[0109] This embodiment illustrates the physical structure of a cylindrical cavity fixture and utilizes a vector network analyzer to measure microwave parameters of capsules containing different drug contents. Due to the similarity between the relative permittivity of air and the relative permittivity of the drug powder inside the capsule, there is significant overlap in the measurement results of S21 and phase parameters. By employing machine learning techniques, this process effectively separates the signal data from a single measurement from the data obtained through multiple measurements, successfully classifying the various drug contents within the capsules.

[0110] Figure 1The cylindrical cavity clamp shown demonstrates its intricate construction details. The main structure is made of copper and plated with a bright metallic coating. Due to the use of a high-frequency band, the surface depth becomes extremely shallow, less than 1 micrometer, much less than the thickness of the metallic coating. Therefore, in use, the clamp material can be considered to be entirely made of metal.

[0111] A network analyzer was used to measure a cylindrical cavity fixture without a capsule, with the aim of comparing the results to simulations. At low frequencies, discrepancies exist between the two results, which can be attributed to the gap between the cap of the plastic tube in the closure and the two orifices. Therefore, the practically usable frequency is lower compared to the frequency calculated for the closure fixture.

[0112] Similar to moisture content, changes in the drug content inside the capsule primarily alter the overall relative permittivity, thereby adjusting the microwave parameters of the clamp. Because the difference in relative permittivity between air and drug powder is relatively small, the changes in the S21 parameter and phase parameter are subtle and difficult to distinguish visually. This embodiment employs machine learning technology to process the collected data, aiming to differentiate between different drug contents exhibiting minute variations.

[0113] Capsules labeled 1 through 5 represent the highest, higher, medium, lower, and lowest drug content, respectively, in descending order. However, the overall results for the S21 parameters and phase in the 12 to 15 GHz frequency range were very similar, making them difficult to distinguish. Therefore, a narrower frequency range (specifically 13 to 13.02 GHz), where relatively large differences exist, was selected for a more detailed examination.

[0114] Because the dielectric constant difference between air and drugs is small, the five different drug contents have little impact on the S21 parameter and phase, making the measurement data difficult to identify. Therefore, machine learning techniques such as Principal Component Analysis (PCA) and Naive Bayes (NB) are used to process and classify the measurement data, aiming to separate the five similar data sets into different groups. Machine learning techniques are used to classify the data. To achieve good separation of capsule measurement results with different drug contents, PCA is first used to extract principal component features from the measurement data of different drug contents to separate the data as much as possible. Then, the five sets of measurement data are labeled into different categories, and two-thirds of them are selected as training data, while minimizing the differences during training and testing. Observing the training and testing process of the model, the NB classifier model can accurately identify the categories of the five different measurement datasets with different drug contents, largely solving the problem of difficulty in manual visual recognition due to small differences in measurement data.

[0115] To minimize the impact of capsule position on the test results, five measurements were performed for each capsule with a different drug content. After removing the capsules with different drug contents, a vernier caliper was slowly pushed to the marked position on the PET tube, and then the measurement was performed using a network analyzer. This process was repeated five times. Based on the results of the five repeated measurements, it was found that each measurement result had a certain influence on the S21 parameter and phase, and this influence occurred in the frequency range overlapping with the analysis frequency range (13.19–13.21 GHz). To improve the reliability of the proposed model, all five sets of measurement data were combined for further modeling analysis.

[0116] Five measurements were performed on capsules with five different drug contents, yielding 25 S21 and phase measurement results. The measurement results for different capsules appeared relatively similar, while measurements for the same capsule showed variation, posing a challenge for content differentiation. A frequency range of 13.19-13.21 GHz was also selected, and corresponding measurement results were obtained. The results show that the distinction between these 25 measurement datasets is not significant, posing difficulties for visual recognition. Therefore, a classifier model based on PCA and an NB classifier model was developed for these 25 measurement datasets. First, PCA features were extracted from these 25 measurement datasets, and then category labels were used to label the measurement data of capsules with five different drug contents. Two-thirds of the data were allocated as training data for the classification model, and the remaining one-third as test data. Finally, an NB classifier model was built using the training data. Certain biases in the five measurements of the same drug content have a significant impact on human identification and judgment. Surprisingly, the NB classifier model achieved a test classification accuracy of up to 94%, indicating that machine learning-based classification models can, to some extent, replace traditional manual visual recognition. The model's classification results were obtained. The results show that the performance of the classifier model built using machine learning techniques is less affected by data bias caused by multiple measurements. The model was able to accurately identify capsules with five different drug contents.

[0117] It is understood that the technical solution presented in this invention provides a novel method for measuring the water content and drug content in capsules using a cylindrical cavity-based microwave testing device. The cylindrical cavity testing method utilizes the relationship between water content and relative permittivity to convert water content into relative permittivity. Simulation studies of capsules with different relative permittivity values ​​revealed that the resonant frequency decreases with increasing water content. This method exhibits a high sensitivity of 17 MHz per percentage water content.

[0118] To verify the device's performance, five capsules with different drug concentrations were selected for testing. Principal component analysis (PCA) and Naive Bayes (NB) machine learning methods were used to analyze the S21 parameters and phase data. This enabled us to achieve 100% separation of different drug concentration scenarios within the frequency range of 13.19–13.21 GHz. To reduce randomness, five iterations were performed, yielding a total of 25 data points, with a separation accuracy of up to 94%.

[0119] Ultimately, the technical solution proposed in this invention can not only distinguish between cases where the relative permittivity difference between the drug and the water inside the capsule is significant, but also effectively distinguish between cases where the relative permittivity difference between the drug and the air inside the capsule is small. This also demonstrates the potential of microwave testing devices for non-destructive and high-speed assessment of drug content in capsules. The device proposed in this study provides a reliable and efficient alternative for assessing drug content in capsules, offering advantages over traditional methods.

[0120] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0121] It should be noted that in the description of this invention, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this invention, unless otherwise stated, "a plurality of" means at least two.

[0122] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0123] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0124] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0125] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0126] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0127] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0128] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A device for measuring the moisture content of an object, characterized in that, include: Measurement module and analysis module; The measurement module includes a cylindrical cavity, a capsule, a transparent plastic tube for the capsule to pass through, two connectors, and two probes; the cylindrical cavity includes a cavity section and a top; the transparent plastic tube passes through the cavity section along the central axis, and the two connectors are located at preset positions on the top, with a probe mounted on each connector; the probes are connected to the analysis module; the capsule contains the object to be measured. When the capsule passes through the transparent plastic tube, the analysis module acquires the insertion loss parameter, analyzes the insertion loss parameter to obtain the water content, including: obtaining the corresponding resonant frequency based on the insertion loss parameter; obtaining the dielectric constant corresponding to the resonant frequency; and obtaining the water content based on the dielectric constant. The analysis module, when deriving the corresponding resonant frequency based on the insertion loss parameter, includes: establishing a recognition classifier model based on the PCA method and the NB classifier model; and using the recognition classifier model to identify the insertion loss parameter to obtain the corresponding resonant frequency. When the analysis module derives the water content based on the dielectric constant, it includes: The water content can be determined based on the dielectric constant using the following correspondence: in, Where is the dielectric constant. MC This refers to the water content.

2. The apparatus according to claim 1, characterized in that, The connectors are symmetrically arranged with respect to the central axis of the cylindrical cavity, and are located on the left and right sides at a quarter position of the central axis.

3. The apparatus according to claim 1, characterized in that, The cylindrical cavity is made of metal and has a diameter of 120 mm and a depth of 23 mm.

Images