提高相关测速中亚时间采样间隔精度的数学处理方法

Mathematic Technique for Improving Velocity Measurement Precision in Sub-Sample Interval by Using Correlation Analysis

该文首先针对相关分析技术测量速度的误差来源进行了分析 ,指出测量误差主要源于采样离散化所引入的误差 ;为了减小测量误差 ,应该对直接计算得到的离散的互相关函数进行进一步处理。利用计算机模拟了穿越两相流光束的透射信号 ,计算了两相流仿真信号的速度 ,并分别给出了未予修正以及采用插值法和最小二乘拟合对互相关函数修正的仿真结果 ;分析表明 ,最小二乘拟合是比较好的修正方法 ,该结论还可以进一步推广到PIV及DPIV测速技术中。最后利用直流电机产生的已知速度场对该文提出的修正方法进行了实验验证。

In this paper the errors, which come form the velocity measurement using the correlation technique, are discussed. It is pointed out that the dispersion (sampling) errors play an important role. In order to mitigate this error, some techniques must be applied to the dispersive cross-correlation function. The signal, which is produced by two parallel light bundles passing through the two-phase flow, is simulated and the flow velocities are calculated by using these simulated data. These dispersive velocity magnitudes are corrected using the interpolation and least square method. These simulation works indicate that the corrected results using least square are better than that using interpolation. An experiment is carried out using a direct current motor and the results coincide with the theoretic analysis.