基于信息熵的粮食干燥塔温度预测控制

Intelligent Control System for Grain Drying Based on Integrated Prediction

针对混流式粮食干燥塔生产过程中,物理化学变化复杂,温度变化呈现的非线性和滞后性,难以准确检测的问题,提出了一种基于信息熵的温度预测方法;首先分别采用支持向量机和灰色预测独立建立第一降速段的温度模型;利用加权方法对两种模型进行集成,最后利用信息熵算法,对加权因子进行优化,提升模型的预测精度;运行结果表明干燥后的稻谷含水量与设定值误差从原来的±24.7%,降低至8.5%,验证本方法在实际生产中的有效性。

For Francis grain drying tower production process, complex physical and chemical changes, temperature changes of the non--linearity and hysteresis difficult to accurately detect the problem, a temperature forecasting method based on information entropy is proposed. First, support vector machine and gray prediction independent establishment of the first spin--down segment temperature model; using a weighting system inte- gration, information entropy algorithm to optimize the weighting factor to enhance the precision of the model, and the actual operating results show that the value of error of dried grain from ±24.7% to reduced to 8. 5%, which verify the validity of the method.