基于高斯函数的风机特性曲线拟合及应用  

Fan characteristic curve fitting based on Gaussian function and its application

作　　者：郭佰灿 乔登攀[1] 范凌云[1] 程国祥 朱继龙 

机构地区：[1]昆明理工大学国土资源工程学院 [2]金川集团有限公司

出　　处：《黄金》2017年第4期28-31,共4页Gold

基　　金：国家自然科学基金项目(51164016);甘肃省科技重大专项计划项目(1203GKDC003)

摘　　要：针对矿山风机特性曲线多项式拟合中存在的曲线振荡和假收敛,拟合的曲线与实际通风机特性曲线的非稳定工作区域形状相差较大等问题,基于最小二乘准则,提出一种用高斯拓展函数拟合风机p-Q曲线的算法。在计算过程中运用MATLAB软件对风机p-Q实测数据进行处理,利用BFS法编程计算了高斯非线性方程组并得到高斯函数各项参数,绘制了高斯拟合结果图,从而给出了求解最佳风机p-Q曲线拟合方程的算法。采用高斯函数可近似拟合包含马鞍形区域与稳定工作区域的风机全压特性曲线,与多项式最小二乘法相比,避免了低次多项式假收敛和高次多项式曲线振荡的问题。In light of the oscillation and false convergence occurring in polynomial fitting of mine fan characteristic curves as well as the notable shape variation of non-stable work areas between the fit curve and actual fan characteristic curve,the paper proposed an algorithm of Gaussian function fitting fan p-Q curves based on least square method. In calculation,MATLAB was employed in processing the measured data of p-Q fan. BFS programmed and calculated Gaussian nonlinear equation sets and obtained the parameters of Gaussian function. The results of Gaussian fitting were charted and the algorithm solving the optimal fan p-Q curve fitting equation was presented. Gaussian function can approximately fit fan＇s full pressure characteristic curve containing saddle-shaped area and stable work area. And compared to the polynomial least square method,the method can avoid the false convergence and oscillation occurring in low order and to high order polynomials respectively.

关 键 词：风机特性曲线 最小二乘准则 高斯函数拟合 BFS法 MATLAB编程 

分 类 号：TD724[矿业工程—矿井通风与安全]