Non-contact detection method for pantograph-catenary contact force based on image processing

Abstract

This paper introduces a non-contact method for detecting contact force between pantograph and catenary (PAC), based on image processing and Kalman filter. A non-contact method for detecting contact force of PAC, which only utilizes the vertical displacement of the collector head, is proposed. First, the images of the collector head area are collected by adding an image acquisition system on the roof of the vehicle. Second, combined with target tracking and template matching, the vertical displacement of the pantograph’s collector head is identified. Finally, based on the Kalman filter considering least squares method and the PAC dynamics model, a method for detecting contact force is constructed. The estimation of the PAC’s dynamic contact force is realized only by using the vertical displacement of the collector head obtained from the image identification. The results of simulation and test-bed results show that the Kalman filter considering least squares significantly has higher accuracy and better robustness compared with the traditional Kalman filter.