Method for identifying the broken and scattered strands of a ground conductor based on a target detection and a long-short-time memory model

Abstract

The invention provides a method for identifying the broken and scattered strands of a ground conductor based on a target detection and a long-short-time memory model, which comprises the following steps: step S1, collecting a ground conductor fault picture in a transmission line as initial training data; step S2, detecting the broken and scattered strands of the ground conductor and marking the processed data; step S3, initializing the pre-trained model, and inputting the data to Faster-Rcnn target detection model for training, training is terminated after the model converges; Step S4, inputting that data extracted from the ground wire in the step S3 into the long-short time memory model in section for train; Step S5: solidifying the long-short time memory model, removing the variables andthe network structure which are only used for training, retaining the weight value of the final detection, and detecting the fault through the solidified model. This method can distinguish the faultline segment from the normal line segment by combining the segmented grounding wire, so the accuracy of identifying broken strand and scattered strand will be greatly improved.

Description

technical field [0001] The invention belongs to the field of electric power inspection fault identification, and in particular relates to a method for identifying broken strands and loose strands of ground wires based on target detection combined with a long-short-term memory model. Background technique [0002] With the growing national economy, the power grid has become larger in order to adapt to its development, and the transmission lines of the power grid have become more and more complex. Manual inspections can no longer meet the current requirements. Therefore, UAV power inspection came into being. After the test of practice, power inspection has become a "coordinated development" of manpower inspection, manned helicopter inspection and UAV inspection. UAVs are flexible and low-cost in terms of line erection traction and line inspection. Problems that cannot be detected by inspection. [0003] The fault detection of UAVs applied to power components is generally dire...

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the existing problem of low identification of fault points of broken strands or loose strands of the ground wire in a complex power transmission environment, and to provide a more general method for conducting ground wires in complex environments. Accurate location of fault points based on target detection combined with long and short-term memory model to identify broken strands and loose strands of ground wire

Images