Spot welding quality control method using artificial neural network using electrode expansion signal as process variable and system

摘要

In order to reduce welding defects and improve productivity by predicting a welding result in a microprocessor-embedded spot welding control device, the present invention mainly uses an electrode expansion signal and an artificial neural network of a welding performing part; An input means for inputting a welding schedule, a welding execution part performing welding according to a welding schedule of the input means, a feed part transferring values of the welding execution part, and amplifying by sensing an electrode expansion signal at an electrode of the welding execution part And a sensing means for filtering and transferring the sensing unit, a database means by dividing the learned artificial neural network parameters by welding conditions, and converting the electrode expansion signal transmitted from the feed unit into a floating-point arithmetic algorithm. The present invention provides a spot welding quality control system using an electrode expansion signal including a central control handler for comparing the result of starting the artificial neural network with the welding strength determined by the welding schedule.