摘要
针对煤层底板突水问题的小样本、非线性特点,采用支持向量回归算法对突水量进行预测,避免了定性分析的局限性。利用遗传算法全局搜索能力的优势,提出了基于遗传算法的支持向量回归参数寻优方法,并建立煤层底板突水量预测的遗传-支持向量回归模型。该模型首先通过遗传算法对训练样本的学习,得到支持向量回归机的最优参数值,然后运用遗传-支持向量回归模型对测试样本进行突水量预测。测试结果表明:与神经网络,传统支持向量回归机的预测值相比,煤层底板突水量预测的遗传-支持向量回归模型精度高,具有较强的泛化能力。
The problem of water inrush from coal floor was characterized by small samples,nonlinear,and using support vector regression algorithm avoided the limitations of qualitative analysis to predict the water inrush quantity.Support vector regression parameters optimization method was proposed based on genetic algorithm using the advantages of the global search capability of the genetic algorithm,and established genetic algorithm-support vector regression model of water inrush quantity prediction from coal floor.First,the model got the optimal support vector regression parameters by genetic algorithm to learn the training samples,and then used genetic algorithm-support vector regression model to predict the water inrush quantity of test samples.The test results show that,compared with the predictive values of neural network and the traditional support vector regression,the genetic algorithm-support vector regression model has higher prediction accuracy and good generalization ability.
出处
《煤炭学报》
EI
CAS
CSCD
北大核心
2011年第12期2097-2101,共5页
Journal of China Coal Society
基金
国家自然科学基金资助项目(60874116)
河北省自然科学基金资助项目(F2010001047)
关键词
煤层底板
突水量预测
遗传算法
支持向量机
支持向量回归机
coal floor
water inrush quantity prediction
genetic algorithm
support vector machine
support vector regression
