Device and method for detecting and calibrating material microwave moisture based on wide spectrum and AI technology

Abstract

The present invention discloses a microwave moisture detection and calibration device and method for materials based on wide spectrum and AI technology, belonging to the field of microwave detection. The present invention adopts a wide spectrum multi-band microwave scanning signal source + ultra-wideband high-gain onboard antenna + digital intermediate frequency digital processing technology, a variety of traditional machine learning + AI combined material moisture content measurement algorithm modeling technology, to achieve the accuracy and stability of high-speed online moisture detection of materials, which is not only applicable to surface moisture, but also to internal moisture, and is not affected by the color and surface structure of the material being tested, water vapor and dust, solving the current microwave moisture detection single frequency point test, low test accuracy, and low test consistency.

Description

Technical Field

[0001] The present invention belongs to the field of microwave detection, and specifically relates to a device and method for detecting and calibrating microwave moisture in materials based on wide spectrum and AI technology. Background Art

[0002] In the online material moisture content detection process in industries such as metallurgy, coal, ore, papermaking, pharmaceuticals, grain, food processing, feed processing, wood processing, biomass power generation, and fiber production, there is an urgent need for moisture detection of biological food raw materials such as mixtures, quartz ore powder, grain, corn, potatoes, and straw.

[0003] Traditionally, the drying method is used to measure moisture. The process involves sampling, sample preparation, drying, weighing and other steps. The workload is large, cumbersome and time-consuming. In the production process, the moisture content of the material cannot be monitored in real time, and it is difficult to meet the real-time moisture control needs of digital workshops and smart factories.

[0004] The infrared moisture detection method is a surface detection technology, which makes it difficult to achieve holographic detection of the overall moisture content of the material being tested. Moreover, factors such as the shape, color, density, and impurities of the material being tested will have a certain impact on the detection accuracy. The instrument is expensive, and the light source needs to be replaced regularly, which requires frequent maintenance and high costs. The lens of the infrared detection method is easily contaminated and easily affected by ambient light, which will affect the accuracy of the moisture test.

[0005] Traditional microwave online moisture testing technology is a single-frequency measurement with a small test range. It is easily affected by material size and temperature, has a single measurement variety, poor accuracy, no IoT interface, and does not meet upgraded testing needs. Summary of the invention

[0006] In view of the above-mentioned technical problems existing in the prior art, the present invention proposes a material microwave moisture detection and calibration device and method based on wide spectrum and AI technology, which has a reasonable design, overcomes the shortcomings of the prior art, and has good effects.

[0007] In order to achieve the above object, the present invention adopts the following technical solution:

[0008] A microwave moisture detection and calibration device for materials based on wide spectrum and AI technology, including a host computer, a main controller, a wide spectrum microwave transceiver analysis integrated module, and an ultra-wideband transceiver antenna;

[0009] The host computer is configured to integrate the moisture prediction model and predict the moisture content of the material in real time;

[0010] A main controller is configured to collect microwave data, issue control instructions and read test data;

[0011] The wide-spectrum microwave transceiver analysis integrated module includes a microwave signal generating circuit and a microwave signal detecting circuit; the microwave signal generating circuit and the microwave signal detecting circuit are respectively connected to the main controller through lines;

[0012] A microwave signal generating circuit is configured to generate an ultra-wideband microwave swept frequency signal;

[0013] The microwave signal detection circuit is configured to compare the amplitude and phase of the ultra-wideband microwave signal after passing through the material to be tested, and obtain the changes in the amplitude and phase of the microwave signal for later data modeling;

[0014] An ultra-wideband transceiver antenna is configured to transmit microwave transceiver signals.

[0015] Preferably, the host computer includes a human-computer interaction interface; wherein,

[0016] The human-computer interaction interface is configured to provide dynamic configuration of parameters including measurement cycle, moisture threshold, density threshold, etc., to display parameter curves including real-time measurement values ​​of material moisture and density, and to view historical data.

[0017] Preferably, the main controller includes an MCU and an FPGA; wherein,

[0018] MCU, configured to write configuration parameters into the registers of the host computer and trigger an enable signal;

[0019] The FPGA is configured to perform digital intermediate frequency signal processing, digital detection, phase detection and microwave channel control.

[0020] Preferably, the device further comprises an ultrasonic sensor and a temperature sensor; the ultrasonic sensor and the temperature sensor are respectively connected to the main controller via lines;

[0021] A temperature sensor is configured to detect the temperature of the material being measured;

[0022] The ultrasonic sensor is configured to detect the thickness of a material being measured.

[0023] Preferably, the ultra-wideband transceiver antenna comprises a transmitting antenna and a receiving antenna; wherein,

[0024] The transmitting antenna is connected to the microwave signal generating circuit and is configured to transmit the microwave signal generated by the microwave signal generating circuit to the measured material and the transmission belt;

[0025] The receiving antenna is connected to the microwave signal detection circuit and is configured to send the received microwave signal to the microwave signal detection circuit.

[0026] In addition, the present invention also mentions a microwave moisture detection and calibration method for materials based on wide spectrum and AI technology. The method adopts the microwave moisture detection and calibration device for materials based on wide spectrum and AI technology as described above, and specifically includes the following steps:

[0027] Step 1: transmitting a microwave signal through a microwave signal generating circuit;

[0028] Step 2: Transmit the microwave signal to the material to be tested and the conveyor belt through the transmitting antenna;

[0029] Step 3: The receiving antenna receives the microwave signal and sends the received microwave signal to the microwave signal detection circuit;

[0030] Step 4: Collect the changes in amplitude and phase of the microwave signal through the microwave signal detection circuit;

[0031] Step 5: Based on the actual moisture content of the material, modeling and algorithm optimization are performed to achieve continuous online testing of material moisture.

[0032] Preferably, in step 5, the specific process of modeling is as follows:

[0033] Step S1: According to the test data, a wide frequency domain sweep is performed on materials with different moisture contents to obtain moisture content and amplitude, and moisture content and phase curves;

[0034] Step S2: constructing an original data set for materials of different thicknesses and different moisture contents through the amplitude and phase data of multiple frequency points;

[0035] Step S3: performing preprocessing operations including filtering and normalization on the original data set, and selecting high-quality measurement frequency bands by recursive feature elimination method;

[0036] Step S4: Through model training and deployment, automatically search for the best regression model type, build a mathematical model, and obtain the moisture content of the material.

[0037] Beneficial technical effects brought by the present invention:

[0038] The present invention adopts wide-spectrum multi-band microwave scanning signal source + ultra-wideband high-gain onboard antenna + digital intermediate frequency digital processing technology, and a variety of traditional machine learning + AI combined material moisture content measurement algorithm modeling technology to achieve the accuracy and stability of high-speed online moisture detection of materials. This is not only applicable to surface moisture, but also to internal moisture. The measurement is not affected by the color and surface structure of the material being tested, water vapor and dust, and solves the current problems of single-frequency point testing of microwave moisture detection, low test accuracy and low test consistency.

[0039] The present invention has high real-time performance and can automatically select the appropriate microwave signal frequency band according to the material. Through machine autonomous learning and intelligent iterative optimization of the model, it provides a fast, high-precision, non-contact, and non-destructive moisture detection intelligent microwave detection system. It will greatly promote the above industries to optimize production processes, enrich high-end manufacturing process equipment systems, improve product quality, speed up production progress, reduce energy loss, reduce carbon emissions, and promote the high-end, intelligent, and green high-quality development of my country's high-end manufacturing industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is a block diagram of a microwave moisture detection and calibration method for materials based on wide spectrum and AI technology of the present invention;

[0041] Figure 2 This is a control flow chart of a microwave moisture detection and calibration method for materials based on wide spectrum and AI technology of the present invention;

[0042] Figure 3 This is a relationship block diagram of the research on microwave moisture detection and calibration algorithm of materials based on wide spectrum and AI technology in the present invention. DETAILED DESCRIPTION

[0043] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0044] like Figure 1 As shown in the figure, the microwave moisture detection and calibration device for materials based on wide spectrum and AI technology consists of a main controller, a wide spectrum microwave transceiver analysis integrated module (microwave signal generation circuit, microwave signal detection circuit), an ultrasonic sensor, a temperature sensor, an ultra-wideband transceiver antenna and other components. The main controller is used to issue control instructions for each module of the whole device and read test data, and is the hardware processing carrier for building mathematical models; the temperature sensor and ultrasonic sensor are used to detect the temperature and thickness of the material being tested. The hardware adds temperature and thickness detection sensors, and the software adds a temperature compensation algorithm to eliminate the influence of excessive changes in temperature and thickness on the test results, making the test results more accurate.

[0045] The wide-spectrum microwave transceiver analysis integrated module of the present invention comprises a microwave signal generating circuit and a microwave signal detecting circuit; the microwave signal generating circuit and the microwave signal detecting circuit are respectively connected to a main controller through lines;

[0046] A microwave signal generating circuit is used to generate an ultra-wideband microwave frequency sweeping signal;

[0047] The microwave signal detection circuit is used to compare the amplitude and phase of the ultra-wideband microwave signal after passing through the material under test, and obtain the changes in the amplitude and phase of the microwave signal for use in later data modeling.

[0048] like Figure 2 As shown, the microwave moisture detection and calibration method of materials based on wide spectrum and AI technology of the present invention also includes a host computer. The host computer integrates a moisture prediction model, predicts the moisture content of the material in real time, controls the whole process of the system operation, and interacts with the user, provides a user-friendly human-computer interaction interface, displays the real-time measurement value and historical measurement value of the moisture content of the material, and can configure the moisture threshold for operations such as material rejection and alarm.

[0049] The core execution process is controlled by the host computer, which controls the entire measurement process, including data acquisition, data preprocessing, moisture prediction, over-limit alarm, data storage, etc. The microwave data acquisition is sent by the host computer to the frequency sweep configuration parameters and start instructions. The entire frequency sweep process is completed in a closed loop in the lower computer, i.e., the main controller, and finally the data is transmitted back to the host computer. The frequency sweep process is completed by the MCU (Microcontroller Unit) and FPGA (Field-Programmable Gate Array) in the main controller. The MCU is responsible for writing the configuration parameters into the register and triggering the enable signal. The FPGA is responsible for receiving and demodulating the IQ signal to obtain the amplitude, phase and other data, and judging whether the data is distorted.

[0050] like Figure 3 As shown, the algorithm research of the present invention mainly includes three parts: data set construction, model training and comparison, and moisture prediction. The microwave amplitude phase and other data are obtained by scanning the single board with different thickness and different moisture content, and the real moisture content of the material is obtained by the oven method to construct the original data set. After the original data is pre-processed by filtering, normalization and other operations, the high-quality measurement frequency band is screened by the recursive feature elimination method to reduce the dimension of the data set and improve the speed. Material moisture prediction belongs to the regression task, and automatic training can be performed to search for the best regression model type, including linear regression model, regression tree, Gaussian process regression model, support vector machine, efficient training linear regression model, kernel approximation model, regression tree integration and neural network regression model. By providing a set of known input data observations (predicted variables) and known responses, supervised machine learning is performed. The model is trained using the observations to generate predicted responses for new input data. The constructed data set is input into the regression learner, different regression models are selected for training, and a model suitable for material moisture detection is selected.

[0051] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the above examples. Changes, modifications, additions or substitutions made by technicians in this technical field within the essential scope of the present invention should also fall within the protection scope of the present invention.

Claims

1. A microwave moisture detection and calibration device for materials based on wide spectrum and AI technology, characterized by: It includes a host computer, a main controller, a wide-spectrum microwave transceiver analysis integrated module and an ultra-wideband transceiver antenna; The host computer is configured to integrate the moisture prediction model and predict the moisture content of the material in real time; A main controller is configured to collect microwave data, issue control instructions and read test data; The wide-spectrum microwave transceiver analysis integrated module includes a microwave signal generating circuit and a microwave signal detecting circuit; the microwave signal generating circuit and the microwave signal detecting circuit are respectively connected to the main controller through lines; A microwave signal generating circuit is configured to generate an ultra-wideband microwave swept frequency signal; The microwave signal detection circuit is configured to compare the amplitude and phase of the ultra-wideband microwave signal after passing through the material to be tested, and obtain the changes in the amplitude and phase of the microwave signal for later data modeling; An ultra-wideband transceiver antenna is configured to transmit microwave transceiver signals.

2. The microwave moisture detection and calibration device for materials based on wide spectrum and AI technology according to claim 1 is characterized in that: The host computer includes a human-computer interaction interface; among them, The human-computer interaction interface is configured to provide dynamic configuration of parameters including measurement cycle, moisture threshold, density threshold, etc., to display parameter curves including real-time measurement values ​​of material moisture and density, and to view historical data.

3. The device for detecting and calibrating material moisture using microwaves based on wide spectrum and AI technology according to claim 1 is characterized in that: The main controller includes MCU and FPGA; among them, MCU, configured to write configuration parameters into the registers of the host computer and trigger an enable signal; The FPGA is configured to perform digital intermediate frequency signal processing, digital detection, phase detection and microwave channel control.

4. The microwave moisture detection and calibration device for materials based on wide spectrum and AI technology according to claim 1 is characterized in that: The device also includes an ultrasonic sensor and a temperature sensor; the ultrasonic sensor and the temperature sensor are respectively connected to the main controller through lines; A temperature sensor is configured to detect the temperature of a material being measured; The ultrasonic sensor is configured to detect the thickness of a material being measured.

5. The device for detecting and calibrating material moisture using microwaves based on wide spectrum and AI technology according to claim 1 is characterized in that: The ultra-wideband transceiver antenna comprises a transmitting antenna and a receiving antenna; wherein, The transmitting antenna is connected to the microwave signal generating circuit and is configured to transmit the microwave signal generated by the microwave signal generating circuit to the measured material and the transmission belt; The receiving antenna is connected to the microwave signal detection circuit and is configured to send the received microwave signal to the microwave signal detection circuit.

6. A microwave moisture detection and calibration method for materials based on wide spectrum and AI technology, characterized by: The microwave moisture detection and calibration device for materials based on wide spectrum and AI technology as claimed in claim 1 specifically comprises the following steps: Step 1: transmitting a microwave signal through a microwave signal generating circuit; Step 2: Transmit the microwave signal to the material to be tested and the conveyor belt through the transmitting antenna; Step 3: The receiving antenna receives the microwave signal and sends the received microwave signal to the microwave signal detection circuit; Step 4: Collect the changes in amplitude and phase of the microwave signal through the microwave signal detection circuit; Step 5: Based on the actual moisture content of the material, modeling and algorithm optimization are performed to achieve continuous online testing of material moisture.

7. The microwave moisture detection and calibration method for materials based on wide spectrum and AI technology according to claim 6 is characterized by: In step 5, the specific process of modeling is as follows: Step S1: According to the test data, a wide frequency domain sweep is performed on materials with different moisture contents to obtain moisture content and amplitude, and moisture content and phase curves; Step S2: constructing an original data set for materials of different thicknesses and different moisture contents through the amplitude and phase data of multiple frequency points; Step S3: performing preprocessing operations including filtering and normalization on the original data set, and selecting high-quality measurement frequency bands by recursive feature elimination method; Step S4: Through model training and deployment, automatically search for the best regression model type, build a mathematical model, and obtain the moisture content of the material.

Images