Research On Intelligence Control Applying To Product Lines Of Gypsum Fiber Board

Gypsum fiber board, as a new type of walling material, has advantages of lightness, highefficiency, fire resistance, damp proof, heat and sound insulation, deciding its widely potentialprospect.The core of gypsum fiber board manufacturing is thickness control. While thicknesscontrol system of gypsum fiber board is a complicated industrial control system, traditionalPID controller's parameters can't reach demanding of excellent property of set value trackingand disturbance restraining at the same time. Therefore, the control effect is unperfect. Aimedat these problems, fuzzy control has already got some compellent achievements, but it alsohas certain problems. Thus, the thesis gives a modified fuzzy control method to design fuzzycontrollers which will be helpful to improve performance of board thickness control systemand product quality of gypsum fiber board.The thesis is based on product line of gypsum fiber board of San Huan Walling MaterialCo. Ltd of Hubei province, through which it gives researches on process flow and thicknesscontrol system, in addition, analyses working principles of continuous press machine andmain factors influencing gypsum fiber board's thickness. Whereas, thickness control systemhas problems of time variance, non-linearity and uncertainty, moreover, it's difficult to decidemembership functions of fuzzy control, which also has shortcoming of large steady state error.After researches about interval valued fuzzy algorithm, the thesis contributes three kinds ofthickness controllers of gypsum fiber board, adopting fuzzy reasoning methods of fuzzymatching interval value, similar measuring interval value and Bidirectional approximateinterval value. And simulation experiments prove the validation of interval valued fuzzyalgorithm. Compared to traditional fuzzy controllers about simulation effects, it shows thatfuzzy controllers based on interval valued fuzzy reasoning have advantages of short adjustingtime and little steady state error. Finally, aimed at three interval valued thickness controllers mentioned above, improvedschemes are put forward about original control laws: adding control points nearby systemsteady point to make the control nearby steady point more exquisite, decreasing system steadyerror and overshoot. Simultaneously, control amplitude away from steady points increases,assuring response speed fast. Simulation proves the reasonableness and efficiency ofimproved scheme. Finally, thickness controller of gypsum fiber board based on Bidirectionalapproximate interval value has optimum effects in all performance points.