粮食收购智能定等系统的研究

Developing a System of Intelligent Grain Grading in Grain Purchase

根据粮食水分和容重确定粮食等级的标准,研究开发了以AT89C51单片机为核心的粮食收购智能定等系统。将模糊神经网络技术应用于该系统中,建构了BP网络─隶属度串联模型。该模型将BP网络的输出作为模糊系统的输入,计算出实测样本属于粮食三个等级的隶属度值,最终输出为测试样本的等级。在虚拟仪器LabVIEW软件平台上开发了系统的工作软件,实现了对水分、容重和等级的一体化检测。实验表明,该系统能够较准确地确定粮食的等级,并具有稳定性好,操作简单,工作界面友好等特点。

Based on the standard of grading grains AT89C51 singlechip as the core, a system of intelligent grain with indexes of grain moisture and test weight, with grading for grain purchase was developed. The following are the working scheme of the system ： Fuzzy neural network technology is applied to the system, and a model of membership degree tandem is established with BP neural network. In this model, all the outputs of BP network are used as the inputs of fuzzy control, and the membership values for three grades of grain sample are calculated ; the final outputs are the grade of tested sample. On the virtual instrument of LabVIEW software platform, a work software is developed, and moisture measurement, test weight measurement, and grain grading are carried out with one kit. The system can grade grains accurately, and exhibits specialities of good stability, simple operation, kind working interface, etc.