电气化公路车辆弓网系统横向偏移智能识别技术

Intelligent Recognition Technology for Lateral Movement of Vehicle Pantograph-catenary System in Electrical Highway

针对当前电气化公路受电弓-接触网系统的弓网接触点检测方法无法适用于电气化公路重型卡车的问题,本文提出了一种基于计算机视觉的检测方法,用于实现对弓网相对位置的智能识别。该方法首先使用BiSeNet v2轻量级语义分割算法,对受电弓和接触网进行分割,通过筛选滤波得到弓网接触点的位置坐标,推算出弓网横向位移、速度值,提取受电弓-接触网系统的横向摆动的空间信息,为横向智能跟踪控制提供闭环输入支持。

In order to solve the problem that the current detection method of pantograph-catenary system is not suitable for heavy trucks on electrified highway,a detection method based on computer vision is proposed to realize intelligent recognition of the relative position of the pantograph-catenary system.The method first uses BiSeNet v2 lightweight semantic segmentation algorithm to segment pantograph and catenary,and obtains the position coordinates of the contact point of the pantograph through filtering and filtering,calculates the transverse displacement and velocity values of the pantograph,and extracted the spatial information of the transverse swing of the pantograph-catenary system to provide closed-loop input support for the horizontal intelligent tracking control.