Signal detection methods and devices in CAN-based communication networks

摘要

PROBLEM TO BE SOLVED: To provide a signal detection method and an apparatus for improving performance in a CAN-based communication network. SOLUTION: The method is a step of receiving a CAN signal from a plurality of CAN communication nodes, a step of preprocessing the received CAN signal in the forward direction based on a forward sampling time point to generate forward input data, and a forward input. Stage of generating forward mask data and forward time interval data based on forward sampling time for data, stage of reverse preprocessing of CAN signal based on reverse sampling time to generate reverse input data, reverse input It includes a step of generating reverse mask data for data and reverse time interval data at the time of reverse sampling, and a step of inputting each of the data into a GRU-D neural network to detect an abnormal signal among CAN signals. [Selection diagram] Fig. 3