论文部分内容阅读
提出基于机器视觉的车辆动态偏移量检测方案,即在被测车辆非共线的4个点上安装高速电荷耦合器件(CCD),随着车辆的运动,CCD高速扫描钢轨,实时分析钢轨相对CCD的运动图像序列,得到各测点相对于钢轨的横向位移和垂向位移,综合各测点的相对偏移量得到整个车体的运行姿态,经进一步分析得到任意点的动态偏移量;研究检测原理,建立由各测点的动态偏移量计算描述车体运行姿态的5个变量和计算任意点动态偏移量的公式,研制模拟车体振动的平台,将所研发的系统与第3方地面测试设备对模拟平台动态偏移量的测试结果进行比较。研究结果表明,采用这2种方法测得相同位置的偏移量波形完全相似,对应幅值相差在-2～2mm以内。
This paper presents a vehicle vision detection scheme based on machine vision. That is, CCD is mounted on the non-collinear 4 points of the vehicle under test. As the vehicle moves, the CCD scans the rail at high speed and analyzes the rail in real time The moving image sequence of CCD is obtained. The lateral displacement and vertical displacement of each measuring point relative to the rail are obtained. The relative displacement of each measuring point is combined to obtain the running attitude of the whole body. The dynamic displacement of any point is obtained by further analysis. The principle of detection is established. Five variables that describe the operational attitude of the vehicle body and the formula to calculate the dynamic offset of any vehicle point are established by the dynamic offset of each measuring point. The platform for simulating the vibration of the vehicle body is developed. 3-party ground test equipment to simulate the dynamic displacement of the platform to compare the test results. The results show that using these two methods to measure the same position of the offset waveforms are completely similar, the corresponding amplitude difference within -2 ~ 2mm.