Development On The Intelligent Controller Of Resistance Spot Welding With Single Chip Microcomputer

The complexity of welding course, melt nuclear uncertainty of forming process, especially some weld course's quality control adopt simple method determine the quality control of spot welding is an very arduous task. High Precision control method must be adopted in the process and quality control of spot welding because the requirement of spot welding joint quality is more and more strict. Furthermore the intelligence control provides a new approach to improve the control accuracy of the spot welding and the quantity of the spot welding. It will become the developing direction. In this thesis, on the basis of fuzzy control and RBF neural networks, the method of spot welding quality control, which integrates constant welding current control with optimizational set function of spot welding technology parameter is presented. A spot welding fuzzy controller is designed, the RBF neural networks optimizational set model of spot welding technology parameter is established and the optimizational control during the spot welding process is realized.The main circuit and its principle are introduced in this thesis. 80C51 microprocessor is used as the core of hardware controller and the hardware circuit which consisted of input circuit, output circuit, sampling enabling signal circuit and alarming are analyzed. Secondly, the controller is introduced. In this thesis, PID controller is first designed and its shortcoming is explained in spot welding control system, so fuzzy controller must be designed. The input and output variables, fuzzy language variables, field, membership function that fuzzy controller deals with, determination of the fuzzy control rule, the methods of fuzzy reasoning, defuzzying and generation of the fuzzy control table are emphatically analyzed . The soft-wares system main program adopting modularization method and subprogram of setting and fuzzy control are generally designed and finished. The simulation experiments are made on control strategy designed.Finally, in allusion to the problem that spot welding technology parameter is set manually and best technology parameter can not obtained in the practicable welding, spot weldingtechnology parameter optimizational set model on the basis of RBF neural networks is designed. It is studied that the technology parameter optimizational set function on the basis of RBF neural networks technology is realized on the single micro-computer system. The optimizational function model about welding time, thickness, welding force, electrode diameter and best welding current is established. Its effectivity is approved through simulation.The spot welding fuzzy controller adopting single chip microcomputer can realize the spot welding constant current control, its dynamic capability is better and speed of its response is fast. Best welding current can accurately be obtained in the spot welding technology parameter optimizational set model on the basis of RBF neural networks and better welding quality can be obtained.