Microwave dielectric sensor capable of realizing independent sensing of double materials and double-point detection method

Abstract

The invention discloses a microwave dielectric sensor capable of realizing double-material independent sensing and a double-point detection method, and belongs to the technical field of microwave sensors. The microwave dielectric sensor comprises a dielectric substrate, and a microstrip transmission line is etched at the central axis position of the top surface of the dielectric substrate; a first split ring resonator and a second split ring resonator are respectively etched on two sides of the microstrip transmission line; the first split ring resonator and the second split ring resonator establish a resonance relation with the microstrip transmission line through magnetic field coupling, and respectively generate independent low-frequency resonance points and high-frequency resonance points in transmission characteristics, and the first split ring resonator is used for sensing the dielectric constant of a first to-be-tested material; and the second split ring resonator is used for sensing the dielectric constant of the second to-be-detected material. According to the microwave dielectric sensor capable of realizing double-material independent perception and the double-point detection method provided by the invention, dielectric constants of two materials can be obtained at the same time in one-time sweep frequency measurement, and frequency crossing and modal interference are avoided.

Description

Technical Field

[0001] This invention relates to the field of microwave sensor technology, and in particular to a microwave dielectric sensor capable of independent sensing of two materials and a dual-point detection method. Background Technology

[0002] In radio frequency and microwave testing scenarios, the dielectric constant of a material is a key parameter characterizing electromagnetic response and energy loss, and it has wide applications in materials testing, industrial quality control, and biomedical analysis. Microwave resonance-based sensors have become a research hotspot in recent years due to their advantages of non-contact operation, high sensitivity, and real-time detection. The basic working principle of these sensors is that when the material under test is placed in a resonant field, it causes changes in the resonant frequency, quality factor, or amplitude. By analyzing these changes, the dielectric constant of the material can be derived. Therefore, developing efficient and accurate dielectric constant measurement techniques is crucial for improving detection efficiency and reliability.

[0003] Existing microwave planar sensors mostly employ split-ring resonators (SRRs) or complementary split-ring resonators (CSRRs) loaded near microstrip lines. For example, Ebrahimi et al. used a pair of identical SRRs to form a differential structure to suppress environmental drift and common-mode interference. However, such devices typically use one side as a reference, making it difficult to achieve independent quantitative sensing of two different materials in a single measurement (DOI: 10.1109 / JSEN.2018.2840691). Kiani et al. used two SRRs of different sizes to achieve dual-frequency operation to improve the measurement coverage and resolution of liquid samples. However, electromagnetic coupling may still exist between the two resonant units, leading to near or unclear resonant zero points, making it difficult to identify multiple points independently (DOI: 10.1016 / j.measurement.2020.107805). While existing dual resonators or differential structures can partially suppress common-mode interference, they typically require one side as a reference, making it difficult to independently and quantitatively sense two different materials in a single measurement. This limits the application potential of such sensors in multi-material synchronous measurement and complex environments. Summary of the Invention

[0004] The purpose of this invention is to provide a microwave dielectric sensor and a dual-point detection method that enable independent sensing of two materials, in order to solve the problems existing in the prior art.

[0005] To achieve the above objectives, the present invention provides a microwave dielectric sensor capable of independent sensing of two materials, including a dielectric substrate, wherein a microstrip transmission line is etched at the central axis position of the top surface of the dielectric substrate, and a first open-loop resonator SRR1 and a second open-loop resonator SRR2 are respectively etched on both sides of the microstrip transmission line. The first open-loop resonator SRR1 and the second open-loop resonator SRR2 establish a resonant relationship with the microstrip transmission line through magnetic field coupling, and generate independent low-frequency resonant points and high-frequency resonant points in the transmission characteristics, respectively. The first open-loop resonator SRR1 is used to sense the dielectric constant of the first material under test, and the second open-loop resonator SRR2 is used to sense the dielectric constant of the second material under test.

[0006] Preferably, the geometry of the first open-loop resonator SRR1 is larger than the geometry of the second open-loop resonator SRR2.

[0007] Preferably, both the first open-loop resonator SRR1 and the second open-loop resonator SRR2 are rectangular ring single-opening structures, with the openings facing away from the microstrip transmission line.

[0008] Preferably, the first open-loop resonator SRR1 and the second open-loop resonator SRR2 maintain a small gap with the edge of the microstrip transmission line.

[0009] Preferably, the dielectric substrate is made of F4B, and the microstrip transmission line, the first open-loop resonator SRR1, and the second open-loop resonator SRR2 are made of metallic good conductor materials.

[0010] A dual-point detection method for microwave dielectric sensors that enables independent sensing of two materials includes: The first test material and the second test material are placed above the first open-loop resonator SRR1 and the second open-loop resonator SRR2, respectively. The transmission characteristics of the microwave dielectric sensor were detected using a vector network analyzer to obtain the low-frequency and high-frequency resonant points. The dielectric constants of the first and second materials under test are calculated using the low-frequency and high-frequency resonant points, respectively.

[0011] Preferably, the dielectric constant is obtained by inversion using an artificial neural network model.

[0012] Preferably, the input to the artificial neural network model is the frequency of the low-frequency resonant point and the high-frequency resonant point, and the output of the artificial neural network model is the relative permittivity of the first test material and the second test material.

[0013] Therefore, the present invention employs the aforementioned microwave dielectric sensor and dual-point detection method capable of independent sensing of two materials, and has the following beneficial effects: (1) By designing open-loop resonator structures of different sizes on both sides of the microstrip transmission line, the spatial separation of the local distribution of electromagnetic field and response frequency is realized, thereby generating two non-interfering resonant zeros in a single measurement, which correspond to the dielectric changes of two material regions respectively. (2) The two resonators have independent frequency responses and are not coupled to each other, which can significantly improve the accuracy and efficiency of multi-point detection, avoid frequency crossover and mode mixing problems, and have the advantages of simple structure, strong decoupling, stable testing and high integration. They are suitable for multi-region material identification and rapid electromagnetic detection scenarios.

[0014] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0015] Figure 1 This is a three-dimensional structural schematic diagram of the microwave dielectric sensor according to an embodiment of the present invention; Figure 2 This is a parameter annotation diagram of an embodiment of the present invention; Figure 3 These are frequency shift comparison diagrams of embodiments of the present invention; (a) shows the transmission characteristic curves when different dielectric constant materials are loaded on top of SRR1 while keeping SRR2 unloaded; (b) shows the transmission characteristic curves when different dielectric constant materials are loaded on top of SRR2 while keeping SRR1 unloaded. Figure 4 This is the equivalent circuit model of an embodiment of the present invention; Figure 5 This is a regression graph of a neural network model according to an embodiment of the present invention; wherein, (a) is the regression effect of the model on the training set; (b) is the regression effect of the model on the validation set; (c) is the regression effect of the model on the test set; and (d) is the regression effect of the model on the entire dataset. Figure 6 This is the error histogram of the neural network model in an embodiment of the present invention; Explanation of reference numerals in the attached figures

[0016] 1. Dielectric substrate; 2. Microstrip transmission line; 3. First open-loop resonator; 4. Second open-loop resonator. Detailed Implementation

[0017] The following detailed description of embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0018] Please see Figures 1-2A microwave dielectric sensor capable of independent sensing of two materials includes a dielectric substrate 1. A microstrip transmission line 2 is etched at the central axis of the top surface of the dielectric substrate 1. A first open-loop resonator 3 (SRR1) and a second open-loop resonator 4 (SRR2) are etched on both sides of the microstrip transmission line 2, respectively. The geometric dimensions of the first open-loop resonator 3 (SRR1) are larger than the geometric dimensions of the second open-loop resonator 4 (SRR2).

[0019] The first open-loop resonator 3 (SRR1) and the second open-loop resonator 4 (SRR2) are both rectangular ring single-opening structures, with the openings facing away from the microstrip transmission line 2.

[0020] The first open-loop resonator 3 (SRR1) and the second open-loop resonator 4 (SRR2) maintain a small gap with the edge of the microstrip transmission line 2.

[0021] The dielectric substrate 1 is made of F4B, and the microstrip transmission line 2, the first open-loop resonator 3 (SRR1) and the second open-loop resonator 4 (SRR2) are made of metallic good conductor materials. Figure 2 The dimensions and descriptions of the sensor are shown in Table 1.

[0022] Table 1 Sensor Dimensions ;

[0023] The first open-loop resonator 3 (SRR1) and the second open-loop resonator 4 (SRR2) establish a resonant relationship with the microstrip transmission line 2 through magnetic field coupling, generating independent low-frequency and high-frequency resonant points in the transmission characteristics, respectively. The first open-loop resonator 3 (SRR1) is used to sense the dielectric constant of the first material under test, and the second open-loop resonator 4 (SRR2) is used to sense the dielectric constant of the second material under test. The microwave signal enters from the input port (SMA connector) and resonates with the resonators on both sides during transmission. 21 Two distinct transmission zeros are formed in the parameters, corresponding to the intrinsic resonant frequencies of SRR1 and SRR2, respectively.

[0024] like Figure 3 As shown, only the material loaded on SRR1 is changed ( When ), the low-frequency null in the transmission response and its second harmonic transmission zero A significant shift occurred, and the high-frequency zero point... The basic structure remains unchanged; similarly, only the material loaded on SRR2 is changed. When ), high-frequency zero point Moving Stable. This independent response characteristic verifies the low coupling between the two SRRs, ensuring selective sensitivity to different materials at different frequencies, thereby enabling independent sensing of two dielectric materials in a single measurement.

[0025] The equivalent circuit model of the sensor structure in this embodiment is as follows: Figure 4 As shown. The parameters of the equivalent circuit model are extracted as follows: ; ; in, It is the resonant angular frequency. It is the -3dB cutoff angular frequency. It is the characteristic impedance of the microstrip line.

[0026] The equivalent circuit model simulates the transmission zeros of each order generated by SRR1 and SRR2 under sensor no-load conditions.

[0027] As the dielectric constant of the loaded material increases, the resonant frequencies of the two SRRs will shift towards lower frequencies. At this point, two types of spectral crossover risks may arise: on the one hand, if... With a significant increase, the low-order resonant point of SRR1 will shift to the left, and its second harmonic 2 will also increase. It will also shift towards lower frequencies, and there are instances where it exceeds the dominant mode resonant frequency of SRR2. The possibility; on the other hand, if Excessive increase, high-frequency resonant point It may also shift towards lower frequencies and approach or cross the low-frequency resonant point. These phenomena will disrupt the operating premise that "low-frequency zeros belong to SRR1 and high-frequency zeros belong to SRR2," impairing the sensor's ability to independently identify the two materials and affecting the clarity of frequency assignment and the accuracy of dielectric constant inversion.

[0028] Therefore, the sensing range is limited in this embodiment. The dielectric constant of the material under test must be limited to 0~15.

[0029] To achieve accurate and rapid inversion of the dielectric constant of two materials, this embodiment employs a regression model based on an artificial neural network (ANN). This model can establish a nonlinear mapping relationship between the resonant frequency response of the sensor under different material loading conditions and the dielectric constant of the material under test, thereby overcoming the computational difficulties caused by the complexity of the frequency response in traditional methods.

[0030] This neural network model is trained using the back-propagation (BP) algorithm, and its network structure includes an input layer, a hidden layer, and an output layer. According to the working principle of this embodiment, the input layer receives two resonant frequencies from the sensor. and As the input vector, where Corresponding to the resonant frequency of SRR1, This corresponds to the resonant frequency of SRR2. The output layer then outputs the relative permittivity of the two materials under test. and .

[0031] To ensure the model's mapping ability and generalization performance, the number of hidden layer nodes was set to 10. During training, normalization was used to preprocess the input data to eliminate the influence of data units and accelerate convergence. The training dataset was divided into training, test, and validation sets in an 8:1:1 ratio for model training, performance evaluation, and preventing overfitting.

[0032] Specifically, during the model training phase, the network continuously adjusts the weights and thresholds using the backpropagation algorithm to minimize the error between the predicted output and the expected output. Once the training process converges or meets the preset performance metrics, an optimized neural network model is obtained. This model can process any set of resonant frequencies measured by sensors. As input, the corresponding dielectric constant can be predicted quickly and accurately. This method effectively solves the problem of simultaneously and independently detecting and quantitatively analyzing two materials within a limited sensing range, greatly improving measurement efficiency and accuracy.

[0033] Figure 5 This demonstrates the regression performance of the neural network model on the training, validation, and test sets, respectively. The coefficient of determination R0 is shown. 2 , is an important metric used to measure the goodness of fit of a neural network model's predictions, and is defined as: ; The regression plot shows that R... 2 All values ​​are greater than 0.99, indicating a good model fit.

[0034] To eliminate the risk of model overfitting or data leakage, linear regression was used to fit data that was completely consistent with the aforementioned neural network model. The results showed that R0... 2 The value of 0.95 further illustrates the stability of the data and that the neural network model does indeed reflect physical laws, eliminating the risk of overfitting or data leakage in the neural network model.

[0035] The neural network model has a high R-value. 2 The value also reflects the stability of the input data, mainly because: , Separately with , The change in frequency exhibits a monotonically decreasing trend, which basically conforms to the physical law of "increased dielectric constant → decreased resonant frequency" in open-loop resonator coupled systems. The coupling between SRR1 and SRR2 is relatively weak. right Mainly sensitive, right It is primarily sensitive to cross-influences, and therefore the functional relationship is relatively close to linear.

[0036] according to Figure 6 The error histogram shown demonstrates that the neural network model exhibits good generalization ability in predicting the relative permittivity. Over 90% of the prediction errors on the training, validation, and test sets are concentrated within the range of 0.1, which fully validates the model's stability and prediction accuracy.

[0037] Therefore, this invention employs the aforementioned microwave dielectric sensor and dual-point detection method capable of independently sensing two materials. Two open-loop resonators with significantly different sizes are loaded on either side of the microstrip transmission line, allowing the low-frequency and high-frequency resonant points to independently sense the relative permittivity of the two materials. Through this heterogeneous size design and weak coupling arrangement, the permittivity of both materials can be obtained simultaneously in a single frequency sweep measurement, avoiding frequency crossover and modal interference. Furthermore, combining a neural network model for data-driven inversion of the dual resonant points not only improves the accuracy and stability of permittivity prediction but also significantly enhances detection efficiency and multi-point testing capabilities.

[0038] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A microwave dielectric sensor that enables dual material independent sensing, characterized by: A medium substrate is etched with a microstrip transmission line at the center axis position of the top surface, and a first split ring resonator SRR1 and a second split ring resonator SRR2 are etched on both sides of the microstrip transmission line respectively. The first split ring resonator SRR1 and the second split ring resonator SRR2 establish a resonant relationship with the microstrip transmission line through magnetic field coupling, and generate independent low-frequency resonance points and high-frequency resonance points in the transmission characteristics respectively, the first split ring resonator SRR1 is used to perceive the dielectric constant of the first material to be measured, and the second split ring resonator SRR2 is used to perceive the dielectric constant of the second material to be measured.

2. The microwave dielectric sensor capable of dual material independent sensing according to claim 1, wherein: The geometric size of the first split ring resonator SRR1 is larger than that of the second split ring resonator SRR2.

3. The microwave dielectric sensor capable of dual material independent sensing according to claim 2, wherein: The first split ring resonator SRR1 and the second split ring resonator SRR2 are both rectangular ring single-opening structures, and the openings are both backward to the microstrip transmission line.

4. The microwave dielectric sensor capable of dual material independent sensing according to claim 3, wherein: A small gap is maintained between the first split ring resonator SRR1, the second split ring resonator SRR2 and the edge of the microstrip transmission line.

5. The dual-material independent sensing microwave dielectric sensor of claim 4, wherein: The material of the medium substrate is F4B, and the materials of the microstrip transmission line, the first split ring resonator SRR1 and the second split ring resonator SRR2 are metal good conductor materials.

6. A two-point detection method of a microwave dielectric sensor capable of independent sensing of two materials, using the microwave dielectric sensor capable of independent sensing of two materials according to any one of claims 1 to 5, characterized in that, Comprising: A first material to be measured and a second material to be measured are placed above the first split ring resonator SRR1 and the second split ring resonator SRR2 respectively; The transmission characteristics of the microwave dielectric sensor are detected by a vector network analyzer to obtain low-frequency resonance points and high-frequency resonance points; The dielectric constants of the first material to be measured and the second material to be measured are calculated respectively by using the low-frequency resonance points and the high-frequency resonance points.

7. The method according to claim 6, wherein the two-point detection method of the microwave dielectric sensor capable of realizing independent sensing of two materials is characterized in that: The dielectric constant is obtained by inversion using an artificial neural network model. 8.The method of claim 7, wherein the method is capable of achieving double-material independent sensing of the microwave dielectric sensor. The input of the artificial neural network model is the frequency of the low-frequency resonance point and the high-frequency resonance point, and the output of the artificial neural network model is the relative dielectric constant of the first material to be measured and the second material to be measured.

Images