Seite - 174 - in Adaptive and Intelligent Temperature Control of Microwave Heating Systems with Multiple Sources

5. ExperimentalResults 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 . 1 ° C ~ 1 . 3 ° C T i m e ( s ) T a r g e tT 1T 2T 3T 4T 5 DT1 = 0 . 6 ° C ~ 1 ° C Figure5.32. SimulationresultsofNNC. The results from the simulations are shown in figure 5.32. From the simulated results, it is clear that the simulated plant is well controlled using the NNC method and the final temperature window is small. Comparing with former presented MPC results, there are one notable difference between the NNC and MPC methods. It is that in the be- ginningof thecontrolprocess, thecontrolledtemperaturesalways lag behind the target temperature. Because the controller has to take a certain time to update its weights from the plant or the NN estima- tor. After the update is finished, the controlled temperatures gradu- ally converge to the target temperature in a faster pace and they track the target temperatureperfectly. The same control behavior also occurred in the real experiments of NNC, as shown by figure 5.33. At both raising-temperature periods, the controlled temperatures have lower temperature increasing rates thantherequiredvalue. Butontheotherhand,thelowerheatingrates leadtoasmalltemperaturewindowinNNCduringthewholeheating process,andthere isnotemperatureovershootat theflat-temperature period. This is the advantage that NNC outperforms the MPC meth- 174