摘要
为了预测爆破振速,讨论了国内外的相关研究方法,提出采用回归型支持向量机(SVMR)进行预测的方法。首先,通过工程实例建立预测模型;其次,使用Matlab仿真了SVMR模型和BP神经网络模型;最后,完成了两种模型下的振动预测。预测结果表明,SVMR方法的性能优于BP神经网络方法。
In order to predict the blasting vibration velocity,related domestic and foreign research methods were discussed,and a new predicted method of using support vector machine for regression(SVMR)was proposed.First of all,prediction models were established based on project examples.Secondly,SVMR model and BP neural network model were created by Matlab software.Finally,experiment simulation under two models had been finished.The predicted results showed that,the performance of SVMR model was better than BP neural network model.
出处
《矿业研究与开发》
CAS
北大核心
2014年第5期108-111,共4页
Mining Research and Development
基金
高等学校博士学科点专项科研基金资助课题(20120143110005)
国家自然科学基金资助项目(51104112
51104111)
中央高校基本科研业务费专项资金资目(2014-IV-140)
武汉理工大学自主创新研究基金项目(2014-JL-007)
关键词
爆破振速
振速预测
回归型支持向量机
BP神经网络
Blasting vibration velocity
Vibration velocity prediction
SVMR
BP neural network