Optimal pole adaptation and control in thermal rapid prototyping applications

This article addresses thermal feedback control of distributed-parameter, dynamic nonlinear heat transfer phenomena involved in rapid prototyping processes. These mechanisms are captured in a numerical simulation of the thermal field generated during such operations. For the design of a multivariable controller, a linearized state-space model of lumped temperatures and heat inputs on a grid is established, and its parameters are identified in-process. The controller gains are adapted in real time through optimization of the closed-loop pole placement by a simplex method, minimizing the distance of the actual from the desired system eigenvalues. This control strategy is validated both by simulation and in the laboratory, where temperature feedback from an infrared camera is used to modulate the power and motion of a plasma-arc torch in laminated object manufacturing