Seite - 170 - in Adaptive and Intelligent Temperature Control of Microwave Heating Systems with Multiple Sources

5. ExperimentalResults Comparing the results in figures 5.30 to the results of the first setup (in figures 5.23 and 5.25), there are a number of obvious distinc- tions. ForthePIDcontroller, thetemperaturewindowsareevenlarger than the results in the first setup, which indicates that the conduction based temperature improvement effect in the second setup is much lower than in the first setup. Unlike the PID controller, the perfor- mance of the linear and the nonlinear MPC methods is both better than the results in the first setup (see figure 5.25), regarding the tem- perature overshoot and the final temperature window aspects. For the linear MPC, it has smaller temperature windows as well as much lower temperature overshoot and oscillations, especially in the high temperature range. This performance fully reflect that without extra disturbances from the thermal conduction, the model constructed in the equation 3.41 is more accurate and close to describe the heating dynamics,whichresults inbettercontrol results. For the nonlinear MPC, the performance improvement is even more obvious than the linear MPC. That is partly because less heating sources are used. At the same time, it profits more from the reduced impact of the thermal conduction. In both MPC methods, the influ- ences fromall coolingeffects includingthethermalconductioncanbe 0 5 0 0 1 0 0 0 1 5 0 0 2 0 0 0 2 5 0 02 0 4 0 6 0 8 0 1 0 0 1 2 0 DT2= 1 6 . 1 °C ~1 6 . 3 °C T i m e ( s ) T a r g e tT 1T 2T 3T 4T 5 DT1= 1 1 °C ~1 1 . 2 °C 0 2 0 4 0 6 0 8 0 1 0 0 (a)PID(newCA3,12sources). 170