Search

KR-20260067621-A - AUTOMOTIVE BODY WELDING PROCESS AUTOMATION SYSTEM USING AI BASED DEEP LEARNING

KR20260067621AKR 20260067621 AKR20260067621 AKR 20260067621AKR-20260067621-A

Abstract

The present invention relates to an automated welding process system for an automobile body, comprising: a welding unit (10) for welding a welding part to the automobile body; an air pressure control unit (20) for controlling the pressure of air supplied to the welding unit (10) to maintain a constant welding quality; a welding inspection unit (30) for verifying the welding state of a part welded to the automobile body using a non-contact induction technology position voltage measurement method; a monitoring unit (40) for managing welding process data to improve quality in the welding process by collecting and monitoring variables that may affect the welding quality of a part welded to the automobile body along with welding process data from the welding unit (10); and a deep learning control unit (50) for managing big data such as welding state inspection data measured by the welding inspection unit (30) and welding process variable data collected by the monitoring unit (40), analyzing the stored big data, and feeding it back to the welding process so that external environmental factors are automatically applied along with the welding process data, thereby enabling automatic quality control of the welding process. By automating the vehicle body welding process, an automated welding process suitable for multi-model, small-batch production can be implemented. Additionally, by learning big data to apply environmental variables that determine welding quality in the automated welding process, the welding quality can be maintained consistently.

Inventors

  • 박종탁

Assignees

  • 유성정밀공업 주식회사

Dates

Publication Date
20260513
Application Date
20241106

Claims (4)

  1. In an automated system for automobile body welding processes, A welding part (10) that allows a welding part to be welded to the vehicle body, and An air pressure control unit (20) that controls the pressure of air supplied to the welding part (10) to manage the welding quality so that it can be maintained consistently, and A welding inspection unit (30) that checks the welding status of a part welded to the vehicle body by a non-contact induction technology position voltage measurement method, and A monitoring unit (40) that manages welding process data to improve quality in the welding process by collecting and monitoring variables that may affect the welding quality of parts welded to the vehicle body along with welding process data by the welding unit (10), and An AI-based deep learning-based automobile body welding process automation system characterized by being composed of a deep learning control unit (50) that manages big data such as welding condition inspection data measured by a welding inspection unit (30) and variable data of the welding process collected by a monitoring unit (40), analyzes the stored big data to provide feedback to the welding process, and automatically applies external environmental factors along with the welding process data so that quality control of the welding process can be performed automatically.
  2. In paragraph 1, The air pressure control unit (20) is, A solenoid coil (21) that generates a magnetic field depending on whether power is supplied, and A magnetic plunger (22) that is operated by a magnetic field generated by a solinoid coil by the supply of power to regulate the pressure of air supplied through a flow path, and A spring (23) that contracts when the magnetic plunger (22) rises due to a magnetic field, thereby supplying air through the air passage, and expands when there is no power supply, thereby causing the magnetic plunger (22) to descend and blocking the supply of air through the air passage. An AI-based deep learning-based automotive body welding process automation system characterized by being composed of a valve seat (24) that opens and closes the flow path in the operation of a magnetic plunger (22) and a spring (23).
  3. In paragraph 1, The monitoring unit (40) is, A welding time checking unit (41) that checks whether the welding of a part is being performed within an allowable range by means of a welding tip installed in the welding part (10), and A current checking unit (42) that checks whether the current applied to the weld part (10) is within an allowable range, and A cooling temperature checking unit (43) that measures and confirms the temperature of a welding tip cooled by cooling water when the temperature of the welding tip detected by a temperature sensor exceeds an allowable range, and A pressure verification unit (44) that measures and verifies the air pressure controlled by the air pressure control unit (20) to weld a part to the vehicle body and to press the part against the vehicle body, and A welding counting unit (45) that counts the number of times welded by the welding unit (10) to ensure uniformity of welding quality, checks the wear condition of the welding tip, and manages the replacement and dressing of the welding tip, and An AI-based deep learning-based automobile body welding process automation system characterized by being composed of an inspection data verification unit (46) that collects welding status data measured by a welding inspection unit (30) and verifies the quality status of the part welding according to the set welding process.
  4. In paragraph 1, The deep learning control unit (50) is, A big data collection unit (51) that receives welding process data collected through a monitoring unit (40) and welding inspection data collected through a welding inspection unit (30) in real time, and A big data storage unit (52) that stores received welding process data and welding inspection data, and A big data processing unit (53) that converts and stores the format of the stored data for data analysis, and A big data analysis unit (54) that analyzes the big data converted through the big data processing unit (53) to check the status of the welding process in real time, and An AI-based deep learning-based automobile body welding process automation system characterized by being composed of an AI judgment unit (55) that automatically determines changes in welding quality according to changing factors by considering real-time status information of the welding process analyzed through a big data analysis unit (54) and external environmental factors of the welding process, and automatically changes variables such as current, pressure, and CYC time applied during welding by the welding unit (10) so that welding quality can be maintained at a constant level.

Description

Automotive Body Welding Process Automation System Using AI-Based Deep Learning The present invention relates to an automotive body welding process automation system using AI-based deep learning for automating the body welding process suitable for multi-product, small-batch production in the automotive manufacturing industry. This invention was supported by the 'Regional Key Industry Development (Regional Specialization Project Support Project) Daegu Regional Specialization Project Support (Robotics)' project. Factory automation involves automating processes within a factory from product design to manufacturing and shipment. By introducing computer systems or industrial robots, it offers advantages such as workforce advancement, quality improvement, and increased productivity. Recently, beyond automatic welding, robotic welding using welding robot systems is being introduced. Since welding robots can complete welding tasks accurately and quickly, they are being applied to various welding methods such as arc welding, spot welding, resistance welding, and TIG welding. Meanwhile, the production process for automobile body brackets still utilizes manual and semi-automated welding due to the diverse range of vehicle models, which is causing many problems such as variations in welding quality and delivery delays. Due to changes in the market environment and customer demands, the domestic automotive industry is undergoing a systemic shift toward multi-model, small-batch production. Consequently, small and medium-sized enterprises (SMEs) are currently developing and applying new concepts of welding and assembly technology incorporating smart, standardized, and automated production methods to reduce initial investment costs and ensure rational factory production line operations. FIGS. 1 and 2 are drawings illustrating the configuration of an automated automobile body welding process using AI-based deep learning according to an embodiment of the present invention. FIG. 3 is a drawing illustrating the detailed configuration of a pressure control unit according to an embodiment of the present invention. FIG. 4 is a drawing illustrating the detailed configuration of a deep learning control unit according to an embodiment of the present invention. FIG. 5 is a drawing illustrating the detailed configuration of a welding variable monitoring unit according to an embodiment of the present invention. Preferred embodiments of the present invention will be described in detail below with reference to the attached drawings. FIGS. 1 and 2 are drawings illustrating the configuration of an automated automobile body welding process using AI-based deep learning according to an embodiment of the present invention. The automobile body welding process automation system using AI-based deep learning of the present invention It is composed of a welding unit (10) that allows a welded part to be welded to a vehicle body, an air pressure control unit (20) that controls the pressure of air supplied to the welding unit (10) to maintain a constant welding quality, a welding inspection unit (30) that checks the welding state of a part welded to a vehicle body by a non-contact induction technology position voltage measurement method, a monitoring unit (40) that manages welding process data to improve quality in the welding process by collecting and monitoring variables that may affect the welding quality of a part welded to a vehicle body along with welding process data from the welding unit (10), and a deep learning control unit (50) that manages big data such as welding state inspection data measured by the welding inspection unit (30) and welding process variable data collected by the monitoring unit (40), analyzes the stored big data to feed back and apply it to the welding process, thereby allowing external environmental factors to be automatically applied along with the welding process data so that quality control of the welding process can be performed automatically. In the present invention, during the process in which welding of a part is performed by the welding part (10), the pressure of the air supplied to the welding part (10) is precisely controlled by the air pressure control part (20), thereby maintaining the welding quality consistently. Referring to Fig. 3, the detailed configuration of the air pressure control unit (20) is explained. The air pressure control unit (20) of the present invention comprises a solenoid coil (21) that generates a magnetic field depending on whether power is supplied or not, a magnetic plunger (22) that is operated by the magnetic field generated by the solenoid coil (21) when power is supplied to regulate the pressure of air supplied through the flow path, a spring (23) that contracts when the magnetic plunger (22) rises due to the magnetic field to supply air through the flow path, and expands when power is not supplied to lower the magnetic plunger (22) to block the supply of air through the flow path,