Liquid detection method, system and device based on multi-frequency resonance

By combining multi-frequency resonance technology with multilayer sensor neural networks, the problems of low accuracy and long detection time in liquid detection are solved, achieving high-precision and fast liquid identification.

CN118980697APending Publication Date: 2024-11-19SUZHOU UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411012025.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-19

AI Technical Summary

Technical Problem

Existing liquid detection methods suffer from low accuracy and long detection times.

Method used

By employing multi-frequency resonance technology combined with multilayer perceptron (MLP) neural networks, the transmission coefficients of liquids over a wide frequency band are obtained to construct training and testing sets. The training set is then used to iteratively train the liquid detection model, which automatically learns and extracts subtle feature differences between liquids, achieving high-precision detection.

Benefits of technology

It achieves high-precision and rapid liquid detection, reducing the time and complexity of manual analysis.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118980697A_ABST
    Figure CN118980697A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of liquid detection, in particular to a liquid detection method, system and device based on multi-frequency resonance, and the detection method comprises the following steps: S1, obtaining transmission coefficients of various liquid samples in a certain frequency band; s2, constructing a liquid detection model, dividing the transmission coefficient into a training set and a test set, and performing iterative training on the liquid detection model by using the training set to extract a transmission coefficient value under a target resonant frequency from the training set to obtain a trained liquid detection model; and S3, detecting the liquid to be detected based on the trained liquid detection model to obtain a detection result. According to the invention, the transmission coefficient characteristics of different liquids can be extracted more accurately, so that high-precision detection of the liquids is realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of liquid detection technology, and in particular to a liquid detection method, system and device based on multi-frequency resonance. Background Art

[0002] In the field of transportation and safety inspection, efficient and non-destructive detection of potentially dangerous suspicious liquids has always been a critical and challenging task. The subtle differences in the molecular polarity and interaction force of liquids show unique electric polarization phenomena under the influence of electric fields, and this characteristic changes dynamically with the fluctuation of the electromagnetic field frequency. It is these subtle physical changes that provide a solid theoretical basis for the accurate identification of liquid types.

[0003] Microwave resonator sensors are used for liquid detection. They have the advantages of high sensitivity, small size, simple structure and low production cost, and can quickly measure the liquid to be tested. However, due to the limitations of liquid detection methods and frequency bands, there is a large uncertainty in the detection process.

[0004] In order to ensure the accuracy and reliability of the test results, the collected data must be carefully processed and analyzed after each measurement. This process not only requires technicians to have deep professional knowledge, but also takes a lot of time and effort to manually build a complex mapping relationship between the liquid type and the key parameters captured by the microwave resonator (such as resonant frequency, signal amplitude, reflection coefficient phase, Q value, etc.). Summary of the invention

[0005] To this end, the technical problem to be solved by the present invention is to overcome the problem of low liquid detection accuracy and long detection time in the prior art. To solve this technical problem, the present invention provides a liquid detection method based on multi-frequency resonance, the liquid detection method comprising:

[0006] S1: Obtain the transmission coefficients of various liquid samples within a certain frequency band;

[0007] S2: constructing a liquid detection model, dividing the transmission coefficient into a training set and a test set, and iteratively training the liquid detection model using the training set to extract the transmission coefficient value at the target resonant frequency from the training set to obtain a trained liquid detection model;

[0008] S3: Based on the trained liquid detection model, the liquid to be detected is detected to obtain a detection result.

[0009] In one embodiment of the present invention, the liquid detection model includes an input layer, multiple hidden layers and an output layer; wherein the input layer is used to receive the training set and convert the training set into array form data; the hidden layer includes multiple hidden neurons, each hidden neuron performs feature extraction on the array form data to obtain a feature value; the output layer converts the feature value into a probability distribution value.

[0010] In one embodiment of the present invention, the characteristic value is calculated as follows:

[0011] Use each hidden neuron to multiply the output of the previous layer or the input of the input layer with the corresponding weight, and add the bias term to get the net input of each hidden neuron;

[0012] The net input is transformed nonlinearly, all negative values ​​are set to zero, and the positive values ​​are kept unchanged, to obtain the output characteristic value of each hidden neuron.

[0013] In one embodiment of the present invention, the net input The calculation formula is:

[0014]

[0015] in, represents the bias term; represents the connection weight from the kth neuron in the (l-1)th layer to the jth neuron in the lth layer; n0 is the number of neurons in each layer.

[0016] In one embodiment of the present invention, the output characteristic value The calculation formula is:

[0017]

[0018] Among them, relu(·) represents the activation function, l represents the layer number, 1≤l≤n, and n represents the total number of hidden layers and output layers.

[0019] In one embodiment of the present invention, the transmission coefficients include S with a frequency range of 2 GHz to 6 GHz and a step size of 0.01 GHz. 21 Transfer curve data.

[0020] Based on the same inventive concept, the present invention also provides a liquid detection system based on multi-frequency resonance, the liquid detection system is used to implement the liquid detection method based on multi-frequency resonance, and the liquid detection system includes the following modules:

[0021] A data acquisition module, the data acquisition module is used to obtain transmission coefficients of multiple liquid samples within a certain frequency band;

[0022] A model building module, wherein the model building module is used to build a liquid detection model, divide the transmission coefficient into a training set and a test set, and iteratively train the liquid detection model using the training set to extract the transmission coefficient value at the target resonant frequency from the training set to obtain a trained liquid detection model;

[0023] A liquid detection module is used to detect the liquid to be detected based on the trained liquid detection model to obtain a detection result.

[0024] The present invention also provides a liquid detection device based on multi-frequency resonance, comprising a multi-resonance liquid sensing unit, a vector network analyzer, a control unit and the liquid detection system based on multi-frequency resonance; wherein the multi-resonance liquid sensing unit is connected to the vector network analyzer, the vector network analyzer is connected to the control unit, and the liquid detection system based on multi-frequency resonance is stored in the control unit.

[0025] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements instructions of the liquid detection method based on multi-frequency resonance when executing the program.

[0026] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the liquid detection method based on multi-frequency resonance are implemented.

[0027] The above technical solution of the present invention has the following advantages compared with the prior art:

[0028] The present invention combines a multi-resonance unit with a multi-layer perceptron (MLP) neural network and applies it to liquid detection. The liquid to be tested is placed in the detection area of ​​the resonant unit, and the transmission coefficient of the liquid is obtained in a wider frequency band. A large amount of data contains rich features. The measured data is trained through a multi-layer perceptron to eliminate the influence of errors. When measuring liquids, it can automatically learn and extract the subtle transmission coefficient characteristic differences between different liquids, thereby achieving high-precision detection of liquids. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to make the content of the present invention more clearly understood, the present invention is further described in detail below according to specific embodiments of the present invention in conjunction with the accompanying drawings, wherein

[0030] Figure 1 is a flow chart of an implementation of a liquid detection method based on multi-frequency resonance provided in an embodiment of the present invention;

[0031] Figure 2 is a schematic diagram of the specific structure of the liquid detection model provided in an embodiment of the present invention;

[0032] Figure 3 is a structural schematic diagram of a liquid detection system based on multi-frequency resonance provided in an embodiment of the present invention;

[0033] Figure 4 is a structural schematic diagram of a liquid detection device based on multi-frequency resonance provided in an embodiment of the present invention;

[0034] Figure 5 is to use Figure 4 The liquid detection device is used to measure water, 100% ethanol, methanol, acetone, ether and 9 ethanol solutions with a concentration range of 10% to 90% (the adjacent concentration difference is 10%), and the transmission coefficient curve is obtained;

[0035] Explanation of the reference numerals in the specification: 100, data acquisition module; 200, model building module; 300, liquid detection module; 10, multi-resonance liquid sensing unit; 20, vector network analyzer; 30, control unit. DETAILED DESCRIPTION

[0036] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments so that those skilled in the art can better understand the present invention and implement it, but the embodiments are not intended to limit the present invention.

[0037] Embodiment 1

[0038] Reference Figure 1 As shown, the present invention provides a liquid detection method based on multi-frequency resonance, and the liquid detection method comprises the following steps:

[0039] S1: Obtain the transmission coefficients of various liquid samples within a certain frequency band;

[0040] S2: constructing a liquid detection model, dividing the transmission coefficient into a training set and a test set, and iteratively training the liquid detection model using the training set to extract the transmission coefficient value at the target resonant frequency from the training set to obtain a trained liquid detection model;

[0041] S3: Based on the trained liquid detection model, the liquid to be detected is detected to obtain a detection result.

[0042] Specifically, Figure 2As shown, the liquid detection model is a multi-layer perceptron, which includes six layers in total, including an input layer, four hidden layers and an output layer; wherein the input layer is used to receive the training set and convert the training set into array form data; the hidden layer includes a plurality of hidden neurons, each hidden neuron extracts features from the array form data to obtain feature values; the output layer converts the feature values ​​into probability distribution values.

[0043] The characteristic value is calculated as follows:

[0044] Use each hidden neuron to multiply the output of the previous layer or the input of the input layer with the corresponding weight, and add the bias term to get the net input of each hidden neuron;

[0045] The net input is transformed nonlinearly, all negative values ​​are set to zero, and the positive values ​​are kept unchanged, to obtain the output characteristic value of each hidden neuron.

[0046] Furthermore, the net input The calculation formula is:

[0047]

[0048] in, represents the bias term; represents the connection weight from the kth neuron in the (l-1)th layer to the jth neuron in the lth layer; n0 is the number of neurons in each layer.

[0049] The net input is transformed nonlinearly, all negative values ​​are set to zero, and the positive values ​​are kept unchanged, to obtain the output eigenvalue of each hidden neuron. Therefore, the output eigenvalue The calculation formula is:

[0050]

[0051] Wherein, relu(·) represents an activation function, l represents a layer number, 1≤l≤n, and n represents the total number of hidden layers and output layers. Preferably, in this embodiment, n is 5.

[0052] The eigenvalues ​​after feature extraction in the four hidden layers are used as the input of the output layer, and the probability distribution value of the eigenvalue is obtained in the output layer

[0053]

[0054] Here, k is the number of neurons in the output layer.

[0055] In this embodiment, the transmission coefficients include S with a frequency range of 2 GHz to 6 GHz and a step size of 0.01 GHz.21 Transfer curve data.

[0056] From the above technical solutions, the present invention uses multiple resonant units, each unit has a unique resonant frequency and frequency response characteristics, to perform a wide-band scan on the liquid placed in its detection area. This design can capture the complex response of the liquid at different frequencies, thereby generating transmission coefficient data containing rich characteristics; the large amount of transmission coefficient data measured is input into the multi-layer perceptron neural network for training. The neural network can automatically learn and extract the subtle transmission coefficient characteristic differences between different liquids through multi-layer nonlinear transformations, thereby achieving high-precision identification of liquid types.

[0057] Embodiment 2

[0058] Based on the same inventive concept as that of the first embodiment, the present invention further provides a liquid detection system based on multi-frequency resonance, and the liquid detection system is used to implement the liquid detection method based on multi-frequency resonance described in the first embodiment. Figure 3 As shown, the liquid detection system includes the following modules:

[0059] A data acquisition module 100, which is used to acquire transmission coefficients of various liquid samples within a certain frequency band;

[0060] A model building module 200, the model building module is used to build a liquid detection model, divide the transmission coefficient into a training set and a test set, use the training set to iteratively train the liquid detection model, so as to extract the transmission coefficient value at the target resonance frequency from the training set, and obtain the trained liquid detection model;

[0061] The liquid detection module 300 is used to detect the liquid to be detected based on the trained liquid detection model to obtain a detection result.

[0062] A liquid detection system based on multi-frequency resonance proposed in this embodiment is used to implement the aforementioned liquid detection method based on multi-frequency resonance. Therefore, the specific implementation method of the liquid detection system based on multi-frequency resonance can be seen from the aforementioned embodiment part of the liquid detection method based on multi-frequency resonance. For example, the data acquisition module 100, the model building module 200 and the liquid detection module 300 are respectively used to implement the steps S1, S2 and S3 of the liquid detection method based on multi-frequency resonance in embodiment one. Therefore, its specific implementation method can refer to the description of the corresponding embodiments of each part. In order to avoid redundancy, it will not be repeated here.

[0063] Embodiment 3

[0064] like Figure 4As shown, the present invention also provides a liquid detection device based on multi-frequency resonance, which includes a multi-resonance liquid sensing unit 10, a vector network analyzer 20, a control unit 30 and a liquid detection system based on multi-frequency resonance as described in Example 2; wherein the multi-resonance liquid sensing unit 10 is connected to the vector network analyzer 20, the vector network analyzer 20 is connected to the control unit 30, and the liquid detection system based on multi-frequency resonance is stored in the control unit 30.

[0065] Furthermore, the multi-resonance liquid sensing unit 10 is a multi-resonance unit liquid sensor.

[0066] When detecting the type of liquid to be tested, the liquid to be tested is placed in a non-metal container, and the non-metal container is placed on the multi-resonance unit liquid sensor (multi-resonance liquid sensing unit 10), ensuring that the bottom of the container completely covers all the resonant units. The multi-resonance unit liquid sensor is connected to the vector network analyzer 20, and different liquids are placed in the liquid placement area to be tested.

[0067] Due to the randomness of the container placement, there will be some differences in the transmission coefficient curve, but this difference is smaller than the difference between different liquids. Sufficient data sets are obtained by placing the container multiple times and measuring.

[0068] The following experiments verify the effectiveness of the liquid detection device of the present invention, using water, 100% ethanol, methanol, acetone, ether and 9 ethanol solutions with a concentration range of 10% to 90% (the adjacent concentration difference is 10%) as the measurement objects. The liquid detection device is used for measurement, and the results are as follows: Figure 5 shown.

[0069] Depend on Figure 5 It can be seen that due to the different types and concentrations of liquids, the transmission coefficient S 21 In the frequency range of 2GHz to 6GHz, different characteristics are shown, and the maximum value, minimum value, average value, slope, curvature, and concavity of the transmission coefficient curve are different in the frequency band. The complex dielectric constants of methanol and ethanol are similar, so S 21 The curves have certain similarities, but have distinguishable features. Liquids with large differences in complex dielectric constants, such as ether and ethanol, show very different features and are easy to distinguish. Therefore, the detection device proposed by the present invention can measure liquids and accurately distinguish them.

[0070] In addition, by Figure 5It can also be seen that the transmission characteristic curves of some of the samples measured have certain similarities. For example, ethanol solutions of adjacent concentrations have similar complex dielectric constants, so it is difficult for the human eye to accurately distinguish their characteristics. However, for liquids with large differences in complex dielectric constants, the characteristics are quite different and easy to distinguish. The trained neural network model can solve this problem well and complete the identification of liquids.

[0071] Embodiment 4

[0072] The present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, instructions for the liquid detection method based on multi-frequency resonance as described in Example 1 are implemented.

[0073] Embodiment 5

[0074] The present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the liquid detection method based on multi-frequency resonance as described in Example 1 are implemented.

[0075] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0076] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0077] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0078] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0079] Obviously, the above embodiments are merely examples for clear explanation and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived from these are still within the protection scope of the invention.

Claims

1. A liquid detection method based on multi-frequency resonance, characterized in that: The following steps are involved: S1: Obtain the transmission coefficients of various liquid samples within a certain frequency band; S2: constructing a liquid detection model, dividing the transmission coefficient into a training set and a test set, and iteratively training the liquid detection model using the training set to extract the transmission coefficient value at the target resonant frequency from the training set to obtain a trained liquid detection model; S3: Based on the trained liquid detection model, the liquid to be detected is detected to obtain a detection result.

2. The liquid detection method based on multi-frequency resonance according to claim 1, characterized in that: The liquid detection model includes an input layer, multiple hidden layers and an output layer; wherein the input layer is used to receive the training set and convert the training set into array-form data; the hidden layer includes multiple hidden neurons, each hidden neuron performs feature extraction on the array-form data to obtain a feature value; the output layer converts the feature value into a probability distribution value.

3. The liquid detection method based on multi-frequency resonance according to claim 2, characterized in that: The calculation method of the characteristic value is: Use each hidden neuron to multiply the output of the previous layer or the input of the input layer with the corresponding weight, and add the bias term to get the net input of each hidden neuron; The net input is transformed nonlinearly, all negative values ​​are set to zero, and the positive values ​​are kept unchanged, to obtain the output characteristic value of each hidden neuron.

4. The liquid detection method based on multi-frequency resonance according to claim 3 is characterized in that: The net input The calculation formula is: in, represents the bias term; represents the connection weight from the kth neuron in the (l-1)th layer to the jth neuron in the lth layer; n0 is the number of neurons in each layer.

5. The liquid detection method based on multi-frequency resonance according to claim 4, characterized in that: The output characteristic value The calculation formula is: Among them, relu(·) represents the activation function, l represents the layer number, 1≤l≤n, and n represents the total number of hidden layers and output layers.

6. The liquid detection method based on multi-frequency resonance according to claim 1, characterized in that: The transmission coefficients include S in the frequency range of 2 GHz to 6 GHz and a step size of 0.01 GHz. 21 Transfer curve data.

7. A liquid detection system based on multi-frequency resonance, characterized in that: The liquid detection system is used to implement the liquid detection method based on multi-frequency resonance according to any one of claims 1 to 6, and the liquid detection system includes the following modules: A data acquisition module, the data acquisition module is used to obtain transmission coefficients of multiple liquid samples within a certain frequency band; A model building module, wherein the model building module is used to build a liquid detection model, divide the transmission coefficient into a training set and a test set, and iteratively train the liquid detection model using the training set to extract the transmission coefficient value at the target resonant frequency from the training set to obtain a trained liquid detection model; A liquid detection module is used to detect the liquid to be detected based on the trained liquid detection model to obtain a detection result.

8. A liquid detection device based on multi-frequency resonance, characterized in that: It includes a multi-resonance liquid sensing unit, a vector network analyzer, a control unit and a liquid detection system based on multi-frequency resonance as described in claim 7; wherein the multi-resonance liquid sensing unit is connected to the vector network analyzer, the vector network analyzer is connected to the control unit, and the liquid detection system based on multi-frequency resonance is stored in the control unit.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the instructions of the liquid detection method based on multi-frequency resonance as described in any one of claims 1 to 6 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the liquid detection method based on multi-frequency resonance as described in any one of claims 1 to 6 are implemented.