接触网受电弓系统的一个新的非接触式故障诊断方法外文翻译、中英对照、英汉互译

接触网受电弓系统的一个新的非接触式故障诊断方法外文翻译、中英对照、英汉互译内容摘要：

R PANTOGRAPH CATENARY SYSTEM To measure automatically the contact strip, the images taken from a camera are used. The main task is to monitor the sequential video frames and detect the anomaly conditions. The obtained images will be analyzed by using canny edge detection algorithm. Edges of pantograph system are extracted for all frames. Contact wire position of each frame is determined from the extracted edge images. The yaxis of this wire continuously changes. This change is saved for each frame. All positions of the contact wire in the video are saved as a signal and then the anomaly of this signal is detected by using statistical parameters. Therefore, the monitoring process will be automated. The first step is to extract the edge of each frame from video. Afterwards, the yposition of the contact wire is detected and added to the YS array. When all of frame is read, the algorithm will produce the position array of yaxis. The features extracted from contact wire signal (YS) are skewness and kurtosis. They can be puted as represent the mean and standard deviation of each window. The edges in an images are the pixels that their brightness’s change suddenly. Edges characterize the boundaries and therefore they have an important topic in image processing . An edge detection algorithm generally consists of three steps: filtering, derivation, and edge detection. In the filtering stage, the image is filtered in order to remove the noise. In the derivation stage, edge areas in the image are brightening up via density change in the edge area. In the last stage, edges of the image are determined. Canny algorithm is known as an optimal edge detection algorithm . Canny algorithm uses 4 Gaussian filter to smooth the image and eliminate the noise in the first step. In the next step, these regions are tracked and the algorithm suppresses any pixel that is not at the maximum. In the hysteresis step, the remaining pixels that have not been suppressed are tracked. In this step, two thresholds are used. If the magnitude is smaller than first threshold, it is taken as nonedge. If the second threshold is smaller than magnitude, this magnitude is taken as an edge. X and y shows the coordinates of the images and σ is the standard deviation. Gn and n show the first derivative of G and direction, respectively. The image is depicted as g. The first step of the algorithm is to apply the canny algorithm to each frame of pantograph’s video. After the edge image is obtained, the position of contact wire will be detected. For each frame, the obtained positions are added to an array. This array will be used to detect the disorder between contact wire and pantograph. The algorithm takes the edge image of the current frame and finds the yaxis position of contact wire. The algorithm starts the searching of contact wire from the first pixel of the current frame. If the algorithm detects a candidate pixel, contact wire method inspect whether the detected pixel is contact wire or not. When the position of contact wire is found, the algorithm will be terminated. For each frame in video, the positions of contact wire is detected and saved to an array. EXPERIMENTAL RESULTS The proposed algorithm is applied to a real pantograph’s video. All images of the pantograph are taken from the roof of the lootive. The video of pantograph system。