Research on the Design of Ceramic Kiln Control System Based on PID

Transcription

1 709 A publcaton of CHEMICAL ENGINEERING TRANSACTIONS VOL. 59, 2017 Guest Edtors: Zhuo Yang, Junje Ba, Jng Pan Copyrght 2017, AIDIC Servz S.r.l. ISBN ; ISSN The Italan Assocaton of Chemcal Engneerng Onlne at DOI: /CET Research on the Desgn of Ceramc Kln Control System Based on PID Je Zhang School of Art Desgn, Luoyang Normal Unversty, Luoyang , Chna Je@163.com In ths paper, the author dscussed the applcaton and desgn of ceramc kln control system based on PID. Ceramc kln plays a very mportant role n the process, and the level of automaton n kln determnes the development of the ceramc ndustry. The frng of ceramc kln s a thermo techncal process wth the varaton of tme, large tme delay and much nterference. The temperature control system s the core of the whole ceramc kln control system. In ths paper, n order to mprove a wde range of temperature control accuracy, by combnng fuzzy and PID control methods a fuzzy PID temperature control system of the thermal analyzer based on ARM mcroprocessor s desgned and mplemented. In ths paper, the hardware has been desgned by usng 32-bt ARM7 processor LPC2368 as the core. The system selected the temperature sensor of Platnum-Rhodum thermocouple. The zero-crossng detecton opt coupler devces and TRIAC make up the mplementaton unt. Human-computer nteracton nterface composes of the key and LCD. 1. Introducton Ceramc kln plays a very mportant role n the process, and the level of automaton n kln determnes the development of the ceramc ndustry. The frng of ceramc kln s a thermo techncal process wth the varaton of tme, large tme delay and much nterference (Jng et al., 2008). It s a broad applcaton to measure and control temperature n the ndustry producton. Especally n some ndustry, such as ol, chemstry, electrc power, metallurgy. The temperature control drectly mpacts on the qualty of the product and ndustral producton process (Dnh and Afzulpurkar, 2007; Yang et al., 2012). Wth the rapd development of the mcroelectroncs technology, the embedded technology and the automatc control theory, the temperature control technology s n the trend of ntellgence development. In ths paper, on the bass of relevant research both at home and abroad, process control for ndustral temperature control system, and amng at the feature of bad expand D/A ablty, the bad dsposal of complex controlled envronment and the nonsupport of fulfllng the real-tme mult-taskng, an ndustral fuzzy PID temperature control system based on ARM s proposed. In hardware, the system adopt ARM9 embedded mcro controller as the man control chp, the temperature detecton module and LCD module have been desgned, and the storage unt has been extended. In addton, n order to download D/Ata, system debuggng and wth the PC machne or devce to communcate more easly, the system also desgn to RS232 seral port crcut, JTAG crcut and Ethernet nterface crcut; In software, the kernel of embedded real-tme operatng system Lnux was cut, confgured, compled and transplant n the fnal. In addton, the programmng of applcatons of temperature detecton module, D/A D/Ata acquston module, LCD module and the control algorthm module has been desgn. The temperature control system wth ths method, overcome the shortage of tradtonal PLC and SCM control, the system has many advantages such as good expand D/A ablty, hgh relablty, fast response, small sze and more ntellgent, etc. In ths paper, the algorthm of temperature control n ndustral producton process has been studed, fnally, fuzzy PID control was chosen as the control algorthm n ths temperature control system. To the boler steam temperature as control object, a smulaton model wth the temperature control system of ndustral boler steam has been establshed. The result show that the fuzzy PID control effectvely mprove the system capacty, such as nonlnear, tme varablty and uncertanty, get better control effect. Please cte ths artcle as: Je Zhang, 2017, Research on the desgn of ceramc kln control system based on pd, Chemcal Engneerng Transactons, 59, DOI: /CET

2 Overvew The mathematcal model of the controlled process could be establshed through analysng ts mechansm accordng to the relaton of materal balance and energy balance wth mathematcal descrpton, ths method has much unversalty. The car spray-pant workshop s dvded nto 2 floors, also ts length s 260m and wh s 50m, buldng area s about 26000m². The frst floor are placed the offce and machne equpment s, on the second floor s the ceramc kln system whch s made up of 5 bg assembled ceramc kln equpment s. The bg assembled ceramc kln equpment s can produce the work demanded ar and send t to all over the workshop. In normal stuaton, 5 assembled needs to send m3/h ar (Jn et al., 2008; Zhu and Wan, 2011; Chen et al., 2009). From the Fg. (1), we can see that when the ceramc kln s workng, the two fans are runnng, outsde fresh ar enters the ceramc kln system through the ar n gate, then nto the frst fltraton system whch can make ar clean, then nto the heat machne whch can heat the ar at a certan value, after that the ar s fltrated agan, then t enters spoutng water system whch can make ar wet to reach a certan humdty, fnally the ar enters the workshop by ar out gate. 3. Modelng of ceramc kln control system In order to demonstrate the effectveness of the proposed mcrocontroller based PID controller a dc-dc converter system was chosen for the practcal mplementaton. The phenomenal advancement n technology has gven rse to modern electronc systems that requre relable, hgh qualty, effcent, small and lghtweght power supply system. Ths has nspred the wde use of dc-dc systems n many ndustral and electrcal systems. The man functon of a DC-DC converter s to convert a fluctuatng DC nput voltage nto a regulated DC output voltage and supply t to a varable-load resstance. DC-DC converters are mostly appled n applcatons such as computers, televson recevers, communcaton devces, medcal nstrumentaton, battery chargers and many other devces that requre regulated DC power. DC-DC converters are also used for DC motor speed control applcatons to provde regulated varable DC voltage. The dcdc converters can be found as step down (buck) or step up (boost) converters. In ths project the buck converter system was chosen for our analyss and demonstraton of the mcrocontroller based PID controller. In order for a buck converter to mantan a constant voltage output, t employs a feedback or closed loop system whch contnuously montors the output voltage and takes correctve acton whenever the output voltage shfts away from the desred value. What ths correctve acton does s to change the duty cycle of the sgnal drvng the MOSFET. Our proposed mcro-controller based PID controller was gong to be that feedback loop whch adjusts the duty cycle of the MOSFET to mantan a constant output voltage on our converter (Mezquta et al., 2014; Sonnemann and Chhay, 2014). Because there are many uncertan factors wthn the room, t's mpossble to correctly determne the mathematcal model of the ceramc kln room temperature through process dentfcaton. The mathematcal model of the controlled process could be establshed through analysng ts mechansm accordng to the relaton of materal balance and energy balance wth mathematcal descrpton, ths method has much unversalty. Because the space of the controlled ceramc kln room s typcally large, the change of temperature and humdty s unreactve, there s strong ant-nterference capablty by tself, and therefore we decded to establsh the mathematcal model of the controlled process through mathematcal dervaton. At the same tme, we know that the temperature and humdty of the room are determned by many factors, such as the temperature and humdty of the outdoor atmosphere, the structure and materal of the external wall, the orentaton of the room, the power of the heatng equpment wthn the room, the quantty of people and the workng nature of the people, and these factors have ther own uncertantes, so t's dffcult to get the accurate mathematcal model of the ceramc kln room temperature (Krmpotc, 2015). The flow characterstc of the regulatng valve s the relaton between the relatve flow of the regulatng valve and the relatve openng of the regulatng valve, and ts functon expresson s (Schnepp et al., 2016; Zhu and L, 2014): max l f ( ) (1) L Relatve flow / max s the rato of the flow under certan openng to the flow under full openng max of the control valve. Relatve openng l/l s the rato of the travel under certan openng l to the travel under full openng L of the control valve. The electrc regulatng valve selected n ths project has equal percentage (logarthm) flow characterstc, the functon expresson of the regulatng valve s:

3 711 l ( -1) L R (2) max Accordng to the model of the electrc valve we selected, R=30. In ths project, the openng of the electrc valve s controlled by the voltage of 0-10V, the control voltage s expressed as ΔU, so expresson (2) s changed to: U 10 max *30 ( t) (3) 30 DN25 electrc regulatng valve s used n ths unt, the heatng capacty of the valve under full openng s 150KW, and the vapour volume flowng through the valve s 0.15m 3 /s, whch s substtuted nto the above expresson (3), the result s U ( t) 0.004*30 10 (4) In ths ceramc kln system, 0.2Mpa vapour s used as the heat source of the heatng ar of the heatng col. Just lke the ordnary heat exchanger, t's consdered that the heat dsspated from the heater s approxmately equal to the heat absorbed by ar flowng through the heater, thus we establsh the followng equaton: 1.2C t rc (5) P K S Where, Cp s specfc heat at constant pressure of ar; Cp=1.01KJ/kg* C ; k s the ar volume flowng through the heater m 3 /s; n ths desgn, the combned ceramc kln system s constant ar volume ar supply system, k=6.5m 3 /s; r s the latent heat of vaporzaton of 0.2Mpa vapour, r=2164 kg; C s the specfc gravty of 0.2Mpa vapour, C=1.651kg/m 3 ; Δt s the temperature rse of the ar after flowng through the heater, n C ; s s the vapour flow entered nto the heater, n m 3 /s Substtute the above known quanttes nto the above expresson (5), the result s: Δt=450s, the result after Laplace transform s : t( s) 450 S (6) The volume of the plant ceramc kln area s V (m 3 ), the ndoor temperature s tn ( C ), the outdoor temperature s tw ( C ), the ar supply temperature s t ( C ), the heat suppled nto the room by the unt s (kj/h), the ar exchange rate of the ceramc kln room s n(1/h); the heat generated by the machnes and persons wthn the plant s m (kj/h); the specfc heat at constant volume s Cv ( kj/ m3*k), the parameter regulated s the ndoor ar temperature tn ( C ). The heat dsspated from the ceramc kln room s out, the heat taken by the return ar s out,a, and the heat permeated through the enclosure s out,b,.e.: out out, a out, b ncvvtn out, b (7) The heat nput nto the ceramc kln room n s the heat nput through ar supply n,n and the heat generated by the machnes and persons wthn the room p, n, nc Vt (8) n n p v p The heat storage capacty of the ndoor ar W s: W C Vt v n (9) Assumng that the enclosure and equpment don't store heat, then the equaton of heat storage capacty change of the ceramc kln room s dw n (10) out Substtute (7), (8), (9) nto (10), the result s (11) n C vv ncvvt p ncvvtn out, b

4 712 For smplfyng calculaton, assumng the enclosure doesn't transfer heat, gnorng the heat generated by persons and equpment, then out,b=0, p=0 substtute t nto (11), then the result s n nt nt (12) n After Laplace transform, the equaton (12) s: tn 1 t s 1 n In the plant, the volume of the area controlled by one ceramc kln s 2800m 3, the ar supply volume of the combned ceramc kln s 50000m 3 /h and the ar exchange rate per second of the room s: n=50000/( )= , substtute t to (13), the result s: 1 tn t (14) 202s 1 The dstance between the ceramc kln area and the unt s 50m, whch are connected va ar duct, the ar speed of the ar suppled by the ceramc kln s 7m/s, then there s about 7s delay of the room's ceramc kln, therefore the transfer functon of the area to be condtoned s 1 7s tn e t (15) 202s 1 Substtute (4) nto (6), then the temperature rse of the unt s: t= U 10, derve (13) U t 1.8*30 10 t w (16) For smplfyng calculaton, assumng the outdoor temperature vares lnearly wth tme tw=13sn((t-6) /12)-7, the orgnal pont of t s Algorthm smulaton of ceramc kln system temperature control Accordng to the transfer functon (15), we establshed the smulaton model as Fgure 1. wth Smulnk of Matlab, f dsturbance sgnal s consdered, the system wll become unstable, tradtonal PID control cannot meet the requrements. In order to stablze the system, we desgned a fuzzy PID adaptve control. Fgure 1: (a) Structure dagram of the assembled ceramc kln and (b) Smulaton model Frst, we should choose the fuzzy varables of Kp, K, Kd, whch are lsted n table 1. Then, we choose fuzzy varables of E and EC, whch are lsted n table 2. An ntellgent adaptve control algorthm s desgned n Matlab to see Fgure 2.

5 713 Fgure 2: PID adaptve control smulaton algorthm Accordng to the actual runnng state of the ceramc kln, we use MATLAB to establsh the fuzzy rules n Fgure 3(a). when the ceramc kln s workng, the two fans are runnng, outsde fresh ar enters the ceramc kln system through the ar n gate, then nto the frst fltraton system whch can make ar clean, then nto the heat machne whch can heat the ar at a certan value, after that the ar s fltrated agan, then t enters spoutng water system whch can make ar wet to reach a certan humdty, fnally the ar enters the workshop by ar out gate. (a) (b) Fgure 3: (a) PID rules and (b) PID adaptve control smulaton char The PID adaptve control smulaton result chart s n Fgure 3 (b). Table 1: Fuzzy varables of Kp, K, Kd Var Z PS PM PB Table 2: Fuzzy varables of E and EC Var NB NS Z PS PB Control engneerng s one of the most mportant part of modern day ndustres. Its man objectve s to avod dsturbances and ensure the desred output n ndustral processes. One of the generc control strateges that s wdely used n ndustral control s the Proportonal Integral and Dervatve (PID) control algorthm. The PID controller has been deemed by some experts to represent the ultmate n control of contnuous processes for whch a specfc mathematcal descrpton (transfer functon) can be wrtten. The PID controller n all ts ablty to elmnate steady-state errors and antcpate the future through ntegral acton and dervatve acton s a

6 714 smple mplementaton of the feedback prncple. More than 95 percent of the control loops n process control are of PID type; a greater number of those loops are actually PI control. PID controllers come n many dfferent forms. They can be mplemented as stand-alone systems n boxes for one or a few loops or as dstrbuted systems for process control. Even systems whch are as dverse as atomc force mcroscopes, cars cruse control systems, or CD and DVD players contan PID controllers. Because they are suffcent for many control problems wth bengn process dynamcs and modest performance requrements, they are found n large quanttes n almost all ndustres. For many years the PID controllers have been mplemented n analog format. After ther ntroducton, these electronc controllers gradually excelled n performance over ther mechancal predecessors, both n terms of performance, speed and cost. As result sngle loop analog controllers have enjoyed a steady popularty and they have reached the hghest level of sophstcaton due to the rapd advances made n the ndustry. But the use of these analog PID controllers however always presented dffcultes n changng the controller parameters whenever there are changes n plant dynamcs due to changes n the operatng condtons. Ths scenaro always requred the engneer to change the hardware n order to change the controller parameters to match wth the new operatng condtons. Durng that tme dgtal computng had not yet advanced and t was very slow and costly compared to the stuaton today. There was lttle software avalable and machne code was used to program the requred soluton to engneerng problems. 5. Conclusons In Chna, when people do actual engneerng projects of factory ceramc kln, due to the nfluence of nput temperature change and other dsturbance, regular PID control method cannot meet the demand of temperature control system. So, n order to solve the temperature control problem, ths paper explored the model of the temperature control system and establshed the model of the ceramc kln. Also, we desgned the fuzzy PID adaptve algorthm and carred out smulaton on the model n Matlab, through the smulaton result n Fg. (5), t could be seen that the control temperature value could reach the set temperature value stably. We have appled the algorthm to the actual control system, the actual temperature control system of the ceramc kln has been runnng for more than 3 years and t s n a good condton. Practce shows that the fuzzy adaptve PID controller has been completely met the system s control demands. Obvously, the ceramc kln model establshed n ths paper s very mportant and useful to study the temperature control of the ceramc kln. Reference Chen J., Cheng M., Xu G., Huang B., L J., 2009, Research on characterstc and ntellgent logc control for temperature process n ceramc roller kln, Journal of Wuhan Unversty of Technology, 31(2), Dnh N.., Afzulpurkar N.V., 2007, Neuro-fuzzy MIMO nonlnear control for ceramc roller kln, Smulaton Modellng Practce and Theory, 15(10), , DOI: /j.smpat Guadarrama J.J., Chavez R.H., 2017, Automatc control of coupled brck klns, Chemcal Engneerng Transactons, 56, , DOI: /CET Jng C., Chun X., Ku X., 2008, Logc Control Algorthm Based on Panboolean Algebra and Its Applcaton for Temperature Control of Ceramc Roller Kln, The Workshop on Computatonal Intellgence & Industral Applcaton IEEE, Krmpotc D., 2015, Informaton mpactogram applcaton for fast detecton of temperature nstablty zone of a two-channel ceramc roller kln, Tehnck Vjesnk, 22(4), , DOI: /TV Mezquta A., Box J., Monfort E., Mallol G., 2014, Energy savng n ceramc tle klns: Coolng gas heat recovery. Appled Thermal Engneerng, 65(1-2), , DOI: /j.applthermaleng Schnepp E., Leonhar R., Korte M., Klett-Drechsel J., 2016, Valdty of archaeomagnetc feld recordng: an expermental pottery kln at Coppengrave, Germany, Geophyscal Journal Internatonal, 205(1), , DOI: /gj/ggw043. Sonnemann T.F., Chhay R., 2014, Non-destructve GPR analyss of threatened Angkoran ceramc klns at Bangkong, Sem Reap, Camboda. Journal of Indo-Pacfc Archaeology, 34, Yang S.., Lu B.L., Pe C.X., 2012, Dstrbuted Control System of Ceramc Materal Kln Based on MVC, Advanced Materals Research, , Zhu Y., Wan J., 2011, Intellgent Temperature Control of Ceramc Kln Based on Data Fuson Technology, Appled Mechancs & Materals, , Zhu Y., L P., 2014, Research on Burnng Zone Detecton Method Based on Flame Image Recognton for Ceramc Roller Kln, Appled Mechancs & Materals, , , DOI: /