Nowadays, single phase AC RSW (Resistance Spot Welding) machines are employed in large scale in industry production. With the appearance of the new production mode, e.g. FMS (Flexible Manufacturing System), the reasonable scheduling is getting more and more important to the system’s performance, including welding quality and the efficiency of the whole manufacturing system. The static, offline and cyclic scheduling methods are used in the traditional production mode of single style and large amount for a long period. While in recent years, to meet the continuously changing requirement of market, the enterprises have to employ the newproduction mode with flexibility, the traditional scheduling methods has problemto suit this change. In this case, there are multiple RSW machines in a workshop and eachof themhas the ability to process all welding operations, work-pieces can be sent to any oneof these machines dynamically. When multiple work-pieces come to welding machines in ashort time, how to schedule the sequence of the welding operation will influence the overall performance of the system. The characteristics of this new production mode are dynamical andmulti-factor involved. While existing scheduling methods normally consider only one factor and make schedule offline.

This thesis proposes a dynamic priority real-time scheduling system for multiple RSW machines, which is based on the principle of MADM (Multiple Attribute Decision Making) theory. The system is implemented based on CAN-bus network, in which IPC (Industrial Personal Computer) and microcontroller are employed as control nodes. Finally, experiments areconducted and the system’s efficiency is proved by the result.