| With the in-depth research on electromagnetic wave application technology at home and abroad,the application of microwave technology in the field of grain quality inspection has attracted much attention.The physical and chemical information carried in the microwave characteristic spectrum of grain is extremely rich,which provides a scientific basis for the application of microwave technology in grain quality inspection.In this paper,a new grain quality and safety testing platform is designed.Taking grain flour as the experimental sample,the electromagnetic wave characteristic spectrum in the microwave band was analyzed and the BP neural network was used to identify and classify it.The fingerprint characteristic spectra of the relevant grains were obtained to provide key technologies that can be used to explore new methods and devices for grain quality testing.The main research work is as follows:(1)The basic principle of measuring the dielectric constant of a substance by the transmission-reflective free-space method is introduced,and the specific method for obtaining it is given.An experimental protocol was designed to verify the feasibility of microwave detection,took wheat flour and talc as samples to conduct dielectric constant measurement experiments and analyze their characteristic spectrum.According to the experimental results: The difference in dielectric constant between wheat flour and talcum powder is obvious under the condition of ensuring the same thickness.that is to say,they can be distinguished based on the difference in dielectric constant of the samples.It shows that the grain quality detection device designed based on microwave technology is feasible in principle,and laid a theoretical foundation for the development and application of grain quality detection system.(2)The hardware design of the detection system.The hardware part of the platform is mainly composed of detection circuit modules,microwave signal receiving and transmitting antennas,and mechanical structure.The detection device takes the STM32 minimum system as the core,which controls and connects the various modules that make up the detection front end.The circuit composition scheme of the detection front end is designed,the functions of each module such as microwave frequency source,mixer,filter,phase amplitude detection and the design flow are also described.Based on the microwave frequency band required by the detection platform,a microwave transceiver antenna with a bandwidth ranging from 0.93 GHz to 9.37 GHz are used.According to the dimensional requirements of the detection front-end circuit and the antenna for receiving and transmitting microwaves,the mechanical structure of the system assembly is designed and manufactured to realize the integrated design of the device.(3)Software development and design of detection system.With the hardware circuit completed,the program is written on the STM32 single-chip microcomputer according to the functions required by each module.The software development of the upper computer is carried out on the Lab VIEW platform,and the grain flour detection program is designed to realize the functions of data transmission and processing,graphic display,and device operation and stop.(4)Construction and testing of microwave inspection platform.After the platform is successfully built,the performance of the detection device is tested through experiments.Five samples of fresh,moldy and aged wheat flour as well as rice flour and talcum powder were tested experimentally,and the results showed that the voltage characteristic curves of all five samples were distinguished significantly.According to the test results,using BP neural network algorithms to build a model for recognition and classification,the identification model predicts up to 95.6%correctly for five samples.It shows that it is feasible to use this device system to detect grain quality,which provides a certain basis for exploring a new grain quality detection platform. |
PDF Full Text Request