Feedforward Control of the Temperature Field in an Experimental Annealing Furnace

Two feedforward control strategies for an experimental annealing furnace equipped with electrically powered infrared (IR) lamps are developed and compared. For the first controller, the optimal time evolution of the electric power supplied to IR-lamps is computed from a dynamic optimization problem. This ensures optimal temperature trajectories and temperature uniformity in the specimen fillet. The second controller is based on an amplified adjustment of the steady-state power distribution obtained from static optimization at an operating point. Here, the temperature uniformity during transients is traded-off against the computing time while still ensuring the best temperature uniformity at the final temperature level. For both strategies, a tailored control-oriented reduced-order model of the 2D spatial temporal temperature evolution is used. The evaluation of the feedforward controllers is carried out with an experimentally validated simulation model. The temperature non-uniformity during transients is less than 0.9 % and reduces to 0.3 % of the setpoint value at the steady state. (C) 2017, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.