Study On Novel Plasma Arc Cutting Technology Based On Pidnn-Fuzzy Controller

Deli Jia,Bo You
2011-01-01
Abstract:Adopting digital inverted plasma arc cutting power as a hardware platform and focusing on its strong nonlinearity and time-varying property, this paper puts forward a variable interval fuzzy-PI double-mode quantification algorithm with a self-adjustable factor in the full domain. The introduction of the PID neural network is for decoupling such parameter variables as the cutting speed and torch height in the multi-parameter hybrid coupling cutting process. This control strategy reduces the complex nonlinear system modeling and realizes real-time and effective online control for the cutting process by combining the advantages of fuzzy control and PID neural network control. Furthermore, the optimized fuzzy control improves the steady-state precision and the dynamic performance of the system simultaneously. The experimental result shows that this control improves the precision, ripples, finish and other comprehensive indexes of the workpiece compared with conventional PI control, and that plasma arc cutting power supply based on the fuzzy-neural network has excellent control performance.
What problem does this paper attempt to address?