Research And Design Of A Grain Moisture Detecting System Based On Capacitance Method

Grain moisture detection is an important and essential link in the food industry. Grain moisture content impacts grains'storing, transporting, and processing etc. At the present stage, grain moisture detection has moved into the direction of intelligence with the rapid development of intelligent sensor technology, computer technology and data processing technology. Various portable detectors and online detecting equipments for grain moisture detection have been developed one after another. However, grain moisture detection has multiple influencing factors, especially the capacitance method, which is most widely applied in domestic. Its detection precision is influenced by grain varieties, volume density and environmental temperature and humidity. This thesis did some research and design work on a grain moisture detection system based on capacitance method after a comprehensive analysis of each influencing factor. The specific research contents and results are as follows:(1) The grain moisture detection experiment was designed and the collecting data was processed and analyzed. The influencing factors on capacitance method were analyzed from three aspects including dielectric properties, physical properties and environmental temperature and humidity. The correlation between dielectric properties, physical properties and moisture content were researched on three types of grain. The impact of environmental temperature and humidity on moisture content detection was analysed. Finally the analysis identified that dielectric property, volume density and environmental temperature were the most significant influencing factors.(2) According to the influencing factors, a grain moisture detecing system was designed based on multi-sensor detection and combination with SCM and PC. Cylindrical capacitance sensor, resistance-strain-chip type weighing sensor, and intelligent temperature sensor were used to detect each parameter in the detecting process and each detection circuit was designed. SCM was used as the core of hardware system to control the detection process.The SCM was treated as inferior computer, and its combination to PC was realized by serial communication. PC realized collecting data from SCM system and doing analyzing and calculating via programming language VB and its control MSComm.(3) Data fusion technology was used in modeling and analyzing the detection data. To build the detection model, methods of quadratic multidimensional regression analysis, BP neural network and parameter estimation fusion were applied. Through compassion of the fitting accuracy of both models, the method of combination of BP neural network and parameter estimation fusion was chosen to build the grain moisture detection model.(4) The hybrid programming of VB and Matlab was applied in the design of the grain moisture detection system based on neural network. VB realized the function of calling Matlab neural network using the technique of ActiveX, and four interfaces of the system were designed including model building, moisture detecting, file managing and system help and corresponding functions was realized.Through debugging of the software and hardware and system testing, the grain moisture detecting system designed could realize the functions of collecting data, communication, modeling, analyzing and calculating, expected design goals reached.