摘要
在建立粮仓储粮重量与粮仓底面关系的基础上,提出了基于支持向量回归的粮仓储粮重量检测模型,给出了基于SVR的粮仓重量检测模型推导。实验表明:本实验选择玉米作为测试粮食类型,总计开展了3次测试,训练SVR检测模型总共得到18个支持向量点,实际预测测试相对误差等于1.4%。以东营粮库作为测试对象,该粮仓总共包含5个粮仓,由稻谷与玉米作为储粮,对各粮仓开展了长期在线测试,各检测点预测误差都在0.4%以内,这表明以SVR为依据的粮仓重量测试模型具备更高的测试精度与鲁棒性,可以满足国家粮仓储粮重量检测的要求。
On the basis of establishing the relationship between the grain weight of granary storage and the grain base of granary,a grain weight detection model based on support vector regression was proposed,and a grain weight detection model based on SVR was derived.The experiment showed that corn was selected as the test grain type in this experiment,and three tests were conducted in total.The SVR detection model was trained to obtain 18 support vector points in total,and the actual prediction test relative error was equal to 1.4%.Dongying grain as test object,the granary contains a total of five granary,from rice and maize as grain storage,online test for a long time on the granary was carried out,all the testing point prediction error within 0.4%,which suggests that based on SVR the granary of the weight test model has higher precision and robustness,can meet the requirements of the state granary storage weight detection.
作者
孟千佳玉
MENG Qian-jia yu(Shandong University of Science and Technology,Dongying 257300 China)
出处
《自动化技术与应用》
2021年第1期110-113,共4页
Techniques of Automation and Applications
关键词
重量监测
压力传感器
检测模型
支持向量回归
检测精度
weight monitoring
pressure sensor
detection model
support vector regression
detection precision
