| Summary: | 博士 === 國立中央大學 === 機械工程研究所 === 92 === Abstract
This thesis describes an automatic welding control system developed for alternating current shield metal arc welding (SMAW). It can replace manual operations which require a well-trained technician. Several methods are applied to control this SMAW welding system.
First, we derive a mathematical model of the welding control system, which consists primarily of an electrode feed-rate mechanism driven by an AC servomotor. Both first-order and second-order dynamics are used to represent this mechanism, when we consider it for an electrode feed-rate velocity control system. The “Matlab IDENTIFICATION Tool Box” is used to estimate the best-fit values of the parameters of the welding control system, using first-order and second-order dynamics, respectively.
A Fuzzy Gain scheduling PID controller can be used to modulate the rate of the first-order electrode feed-rate mechanism that regulates the arc current. The electrode feed-rate mechanism for this controller is driven by an AC servomotor which can both compensate for the molted part of the electrode and for the undesirable fluctuations of the arc length during welding operation. It can also be easily applied to any welding system where the electrode is consumed during the welding process. Furthermore, an adaptive sliding mode controller can estimate the uncertainty bounds, and modulate the rate of the electrode feed mechanism regulating the arc current. We also examine the performance of this adaptive sliding mode controller experimentally. The simulations and the experiments both show that the sliding mode occurs.
The sliding surface can be used to as antecedent part to simplify the inference rule base for the fuzzy logic control (FLC). An adaptive fuzzy sliding mode controller uses an adaptive law to update the consequent part of the FLC to approximate the equivalent control of a conventional sliding mode control, modulating the rate of the electrode feeding mechanism that regulates the arc current. This is one method that we have applied to control the SMAW system. The stability of the arc welding current, the uniformity of the welding beads, and the reduction of spatter generation all can be dramatically improved by using this control method.
Finally, in this thesis, we introduce an adaptive fuzzy logical controller design using a single fuzzy input variable signed distance . Based on the similarity between the control rule table of the prevalent FLCs and an ordinary SMC with a boundary layer, the single input variable signed distance can be used to reduce the number of fuzzy reasoning rules, compared to the conventional two-dimensional FLCs. Again, an adaptive law is used to obtain the FLC parameters, and is applied to approximate the equivalent control part of the SMC, so that the system states can be forced to zero, thus guaranteeing the stability of the adaptive fuzzy signed distance sliding mode control. The feasibility and the simplicity of this controller for an SMAW system is verified by a simulation. The results show that the expected approximation of the sliding property occurs. The experimental results verify the feasibility of the proposed controller and also show that it can effectively perform arc welding, as well as effectively initiate the welding arc.
|
|---|
