传统的最小二乘法拟合流速剖面存在一定的困难.利用霍夫变换抗干扰(或噪声)能力强的优点,可将干扰点(或噪声)剔除,然后用霍夫变换拟合流速剖面.当精度要求较高时,可将各拟合点到拟合直线的距离作为判别标准,对经霍夫变换剔除干扰点(或噪声)后的数据点再进行最小二乘法拟合.经过检验与应用表明,霍夫变换拟合的流速剖面比传统的最小二乘法更接近实测值.
It is difficult to simulate a velocity profile with Hough Transform to simulate a velocity profile by using the the traditional least square method. This paper takes the anti-noise advantage of Hough Transform to eliminate the noise points. If a higher accuracy resolution is required, the distances of points to the fitting line can be taken as the judgment standard, and the data eliminated noise points by Hough Transform can be simulated by the least square method. Results of inspection and application show that the velocity profile simulated by Hough Transform is more close to the measured values than that simulated by the least square method.