当前位置: 首页 > 文章 > 基于ε-SVR的风筛式清选装置清选性能预测研究 农机化研究 2018,40 (4) 26-30+36
Position: Home > Articles > Prediction for Working Performance of Air-and-screen Cleaning Unit Based on the ε-SVR Method Journal of Agricultural Mechanization Research 2018,40 (4) 26-30+36

基于ε-SVR的风筛式清选装置清选性能预测研究

作  者:
梁振伟;李耀明;周全;马征;魏纯才;王建鹏
单  位:
江苏大学现代农业装备与技术教育部重点实验室
关键词:
清选;联合收获机;风筛式;ε-SVR
摘  要:
在分析传统预测模型不足之处的基础上,为了能方便地预测清选参数对清选性能的影响,将一种支持向量机模型引入风筛式清选装置的清选性能预测领域,探讨了样本容量大小对ε-SVR回归模型预测性能的影响,并与BP预测模型进行了对比。分析结果表明:采用非启发式Grid Search方法及启发式GA和PSO方法寻求ε-SVR模型最佳参数,可避免凭经验选取参数的随机性,在具有小样本的清选性能预测中,ε-SVR模型预测性能优于BP模型。
译  名:
Prediction for Working Performance of Air-and-screen Cleaning Unit Based on the ε-SVR Method
作  者:
Liang Zhenwei;Li Yaoming;Zhou Quan;Ma Zheng;Wei Chuncai;Wang Jianpeng;Key Laboratory of Modern Agricultural Equipment and Technology,Ministry of Education,Jiangsu University;
关键词:
cleaning;;combine harrester;;air screen;;ε-SVR
摘  要:
On the basis of analyzing disadvantages of conventional prediction model of air-and-screen cleaning device,a new regression model based on support vector machine was proposed to predict and control of cleaning process precisely.Parameters ofε-SVR models were determined utilizing non-heuristic Grid Search、heuristic GA and PSO which could avoidthe choice of randomness.The effect of samples in different on prediction performance ofε-SVR was analyzed compared with BP.The results indicate that the prediction property ofε-SVR is better than BP especially for small sample space.

相似文章

计量
文章访问数: 18
HTML全文浏览量: 0
PDF下载量: 0

所属期刊

推荐期刊