主通风机风压特性方程求解算法研究  

Research on algorithm of the main ventilator wind pressure characteristic equation

作　　者：宫良伟[1,2] 

机构地区：[1]中国矿业大学矿业工程学院,江苏徐州221116 [2]皖北煤电集团,安徽宿州234000

出　　处：《煤炭学报》2012年第5期868-873,共6页Journal of China Coal Society

摘　　要：为解决高次多项式拟合风压特性曲线产生的病态性问题及确定风压特性曲线最佳拟合次数,提出使用正交多项式处理,可避免病态方程组;并提出均方误差δ不能作为判别拟合效果的准则,而在误差符合零平均正态分布时,可使用标准差σ作为判断拟合效果的标准。基于正交多项式、最小二乘原理及标准差σ,给出了求解最佳风压特性曲线拟合方程的算法,并得出风压特性曲线5次或6次不是共同的最佳拟合次数,最佳拟合次数因通风机不同而异。To solve the morbidity produced by high order polynomials fitting wind pressure characteristic curve and determine the optimal power of data fitting to characteristic curve of the fan,orthogonal polynomial in least squares method was used to avoid morbid equations,and MSE δ（mean-square error）cannot be served as the criterion of judgment on fitting effect,and when error accords with zero average normal distribution,SD（standard deviation） σ can be the standard of fitting effect.Based on the orthogonal polynomial,least-square principle and SD σ,the algorithm of optimal wind pressure characteristic curve fitting equation were given,and research conclusions about wind pressure characteristic curve are that 5 or 6 order polynomials is not common optimal power of data fitting to characteristic curve of the fan,and the optimal power is different with the fan itself.

关 键 词：主通风机 风压特性曲线 最小二乘拟合 病态方程组 正交多项式 均方误差δ 标准差σ 

分 类 号：TD441[矿业工程—矿山机电]