Intelligent control of laser-based manufacturing processes and systems

Lasers have been incorporated into manufacturing as an alternative to conventional approaches to improve product quality as well as to reduce production cost. The success of lasers is partly due to their high beam quality that is a direct result of the condition of components involved in the lasing process. In practice, maintaining high beam quality typically requires frequent and time consuming intervention by skilled human operators for inspection, servicing, and reassembling laser components. These components may be indirectly monitored to make those inspections fast and effective; consequently, maintenance downtime is minimized and productivity of man and machine is improved.; The objective of this research was to unify process control, condition monitoring, lifetime prediction, and condition-based maintenance scheduling for laser-based manufacturing processes. This unification is expected to allow manufacturers to safely and continuously operate their lasers without causing damage to either work-in-process parts or the laser systems themselves. Consequently, the manufacturers improve the productivity as well as avoid a long machine downtime.; Among laser-based processes, free-forming was selected as an illustrative case for the development of an intelligent control. Free-forming gives manufacturers an alternative of part producing by adding material onto parts rather than the conventional method of subtracting material from parts. A real-time process control is necessary to avoid the costs of continuing to produce unacceptable products. For the laser free-form process, a complex, intensively computational, numerical model exists. A metamodel, developed to represent this model, was used for controller design. A non-identifier-based adaptive process controller was then developed. Using this, a condition-based maintenance controller was developed to assist the process controller to quickly determine the laser condition. Based on the current laser condition: the maintenance controller estimates the remaining useful lifetime of its laser; recommends the appropriate laser power to the process controller; interacts with a maintenance crew agent to create a condition-based maintenance resource schedule.; The performance of this intelligent controller could be improved in the future by the use of computing technology such as FPGAs, which would make possible real time calculations for those cases where models cannot be simplified, thus permitting a real-time control capability.