基于图像识别技术的随机线缆束建模及分布参数统计分析

Modeling and statistical analysis of distribution parameters of random cable bundles based on image recognition technology

  • 摘要: 提出了一种对实际弯曲随机捆扎线束的建模方法,该方法首先基于图像识别技术,使用实际线束在侧视和俯视方向的两幅照片来实现弯曲线束轴心三维坐标的重建,然后再基于随机转移路径方法实现弯曲线束的捆扎随机性。基于该建模方法,通过蒙特卡洛模拟来分析弯曲随机线束分布参数的统计特征,发现自电感、互电感和互电容均值沿线变化趋势与线束高度变化趋势一致,自电容均值则趋势相反;自电容、自电感和互电感的变异系数与线束高度存在负相关特征;捆扎随机性不会改变自电感和自电容均值,但是会降低互电容与互电感均值。

     

    Abstract: In this paper, a modeling method of actual bending random bundled wire harness is proposed. Firstly, based on image recognition technology, the three-dimensional coordinates of bending wire harness axis are reconstructed by using two photos of actual wire harness in side view and top view; then the random bundled wire harness is realized based on random transfer path method. Based on this modeling method, this paper analyzes the statistical characteristics of distribution parameters of bending random wire harness by Monte Carlo simulation, and finds that the variation trend of self inductance, mutual inductance and mutual capacitance along the line is consistent with the variation trend of wire harness height, while the trend of self capacitance is opposite; the coefficient of variation of self capacitance, self inductance and mutual inductance has negative correlation with wire harness height; the bundling randomness is not obvious It will change the mean value of self inductance and self capacitance, but reduce the mean value of mutual capacitance and mutual inductance.

     

/

返回文章
返回