Design of grain moisture detection system based on microwave

The moisture content of grain is a key indicator of grain quality, which directly affects the grain procurement, transportation, storage, processing, trade and other processes. At present, when domestic grain is purchased, the moisture content of grain is judged by touching or biting with hands or traditional detection methods, which has problems such as extremely unreliable measurement results, long detection time, and waste of manpower and material resources. In order to quickly and accurately detect the moisture content of grain, a microwave-based grain moisture detection system is designed. By detecting the changes in microwave amplitude and phase before and after the microwave signal interacts with the measured grain, the moisture content of grain is calculated.

1 Microwave moisture detection
Microwave moisture detection is a new non-destructive detection technology developed in recent years. It has the advantages of high detection accuracy, wide measurement range, good stability, easy dynamic detection, low sensitivity to the environment, and can be carried out under relatively harsh environmental conditions. Microwaves, as a very high frequency electromagnetic wave, have strong penetration. It detects not only the moisture on the surface of grain, but also the moisture content inside the grain without damage. The dielectric constant and attenuation factor of water in grain are much higher than the dielectric property values ​​of dry matter, and water as a polar molecule is polarized under the action of microwave field, showing special sensitivity to microwaves. Microwave grain moisture detection uses the absorption and reflection of microwave energy by water to cause changes in microwave signal phase, amplitude and other parameters to detect moisture content. Microwave moisture detection is gradually replacing traditional moisture detection methods such as capacitance method and resistance method with low accuracy, high sampling requirements and poor adaptability, and has become an ideal grain moisture detection technology. Microwave moisture detection can adopt transmission and reflection detection methods, and its microwave sensor layout is shown in Figure 1.

Generally, when the material thickness is relatively thin, the transmission detection method is used; when the material thickness is relatively thick and the density is relatively high, the reflection detection method is used. Microwave detection is a depth measurement technology, and the measured result is representative of the total volume moisture, which is much better than the surface measurement technology.

2 System Design
2.1 System Structure Design
Based on theoretical analysis and a large number of experiments, a microwave-based grain moisture detection system is designed, as shown in Figure 2. The system is mainly composed of a microwave generator, a microwave sensor antenna, a temperature sensor, a detection controller, and analysis and processing.

The microwave generator operates at a frequency of 10.5 GHz, and the microwave sensor is arranged in a transmission detection method. The isolator allows the forward-transmitting microwave to pass through without attenuation or with very little attenuation, while the reverse-transmitting microwave has a greater attenuation. Using the isolator, the reflection caused by load mismatch can be absorbed by the isolator and cannot return to the signal source, so that the signal source can work stably. The detector converts the microwave signal into an electrical signal. After amplification, filtering and A/D conversion by the detection controller, the detection controller communicates with the computer through the serial bus. The computer can complete data analysis and real-time display. The detection controller can set system parameters, calibrate moisture, display detection results, etc. Temperature compensation is performed through the temperature sensor signal to obtain the ideal linear relationship between the microwave detection signal and the moisture content of the grain, thereby improving the detection accuracy of the system.
2.2 Hardware Design
The grain moisture detection controller is the core part of the hardware design. It consists of an amplification and filtering circuit, A/D conversion, a microcontroller, a keyboard, an LCD display, and a serial bus interface, as shown in Figure 3. The microwave electrical signal picked up by the microwave sensor probe is amplified, filtered, and then A/D converted. The grain moisture estimation result is displayed in real time on the LCD. The keyboard interface can be used for controller parameter setting, moisture calibration, etc. Data communication with the computer is carried out through the serial port, and the CAN bus is a cascade interface for multiple controllers.

[page]

The microwave sensor analog signal processing determines the moisture detection range and detection accuracy of the entire system. Figure 4 shows the sensor signal processing circuit. The microcontroller uses Microchip's PICl-8F6527. The processor uses nanowatt technology, low power consumption, strong anti-interference ability, and rich peripheral interfaces, such as CCP module (PWM), MSSP module (SPI, I2C), EUSART module, A/D conversion module, etc., which can meet the application requirements of the system. 4 crystal oscillator modes, up to 40 MHz. There is a watchdog circuit inside, which can be programmed online serially (ICSP). The A/D converter uses AD7806, 16-bit sampling, and its reference voltage source 3 V is provided by AD780. The A/D sampling resolution is 45.8μV/bit, which makes the system have a higher detection accuracy. AD7806 has a sampling self-correction function to ensure the accuracy of sampling, and communicates data with the microcontroller through the SPI bus. After the power-on initialization is completed, the AD7806 provides an interrupt signal to the microcontroller through the pin from high to low, notifying the microcontroller that a new A/D conversion data is ready and can be read. The amplification factor of the signal processing circuit can be adjusted online, which is completed by controlling the digital potentiometer AD5227 through PICl8F6527. The two are connected by 3 data lines, which improves the moisture detection range of the system. The entire system adopts 5 V working voltage and low power consumption design. The interface with external devices is photoelectrically isolated to reduce external interference and improve the reliability of the system.
2.3 Software Design
The grain moisture detection system software consists of data acquisition, moisture value calibration, moisture value calculation, system sensitivity adjustment and display module. The system sensitivity adjustment module can adjust the analog signal amplification factor according to the sampled data to improve the robustness of the system. Figure 5 is the real-time microwave moisture detection software interface.

Figure 6 shows the software design process. When the detection system is installed for the first time or the material variety to be detected changes, the system moisture calibration and parameter setting are required. Generally, 2 or more data points need to be calibrated, and data fitting processing is performed according to the calibration value. After the system is initialized, the data acquisition module collects a small section of microwave signal and smoothes this section of data. The bubbling method is used to sort first, and the middle data is selected for weighted average, and the sampled data is temperature compensated. The moisture value estimation module calculates the grain moisture content value through the linear matching algorithm based on the prior calibration value and the result after smoothing preprocessing and displays it in real time.

3 Results and analysis
1) Experimental materials The experimental samples are 40 kg each of 8 different moisture levels of wheat (moisture range 8% to 20%). The materials required for the standard drying method include an electric constant temperature oven, an electronic balance with an accuracy of 0.001 g, an electric crusher, an aluminum box, etc.
2) Methods The wheat is measured on the test platform using a microwave-based grain moisture detection system to obtain the moisture value of the wheat. At the same time, a certain sample is taken and the standard moisture value of the wheat is obtained using the standard drying method. The system measured values ​​were compared and analyzed with the standard values ​​obtained by the drying method. The experimental data are shown in Table 1.

The grain moisture detection system has been tested experimentally to have a moisture content range of 8% to 20% and a measurement accuracy of 0.5%. A large number of on-site experimental tests have shown that the system can fully meet the needs of grain moisture content detection.

4 Conclusions
The grain moisture detection system based on microwaves can continuously and accurately detect the moisture content of grains, providing a strong technical guarantee for the purchase, transportation and storage of grains. Compared with traditional grain moisture detection methods such as capacitance and resistance, microwave detection has fast speed, high accuracy and good stability, which solves the current problems of long detection time, extremely unreliable measurement results and inability to realize online detection during domestic grain purchase. A large number of indoor and outdoor experiments have shown that the system can meet the needs of moisture content detection in the process of grain purchase, storage and processing, and has broad application prospects and economic benefits.