Server for predicting and diagnosing of failure of electric vehicle entrance door, System for predicting and diagnosing of failure of electric vehicle entrance door and Method for predicting and diagnosing of failure of electric vehicle entrance door

摘要

According to an aspect of the present invention, there is provided a method comprising: collecting input data for each state including current, voltage, operation time, and movement position of a door of a door motor in a plurality of states by a data collecting device (step S1); Time-based data including the voltage and current of the motor according to the operating time of the door motor for each state is converted into position reference data including the voltage and current of the motor according to the moving position of the door by the door failure diagnosis and prediction server becoming a step (step S2); calculating a plurality of mathematical feature variables by substituting the location reference data for each state into a plurality of mathematical feature extraction functions by the door failure diagnosis and prediction server (step S3); performing principal component analysis on the plurality of mathematical feature variables by the door failure diagnosis and prediction server to extract N principal components (step S4); Building a failure diagnosis model capable of diagnosing one of the plurality of states by a failure diagnosis algorithm for a specific location section based on the N main components by the door failure diagnosis and prediction server (step S5) ; and diagnostic input data including current, voltage, operation time, and door movement position of the door motor from the data collection device, the diagnostic input data is input to the failure diagnosis model, and the door is one of the plurality of states A method for diagnosing and predicting a door failure can be provided, including a step (step S6) in which it is determined that the state of