Jounal of Theoetical and Applied Infomation Technology 28 th Febuay 2013. Vol. 48 No.3 2005-2013 JATIT & LLS. All ights eseved. ISSN: 1992-8645 www.jatit.og E-ISSN: 1817-3195 HYBRID FUZZY PD CONTROL OF TEMPERATURE OF COLD STORAGE WITH PLC 1 DING WANG, 2 YUAN FANG, 3 MINGYUE MA, 4 FENGHUA QI 1 Automatic Contol Key Laboatoy of Heilongjiang Univesity, Heilongjiang Univesity, Heilongjiang, P. R. China 2 Habin Radio and TV Univesity, Heilongjiang, P. R. China 3 Automatic Contol Key Laboatoy of Heilongjiang Univesity, Heilongjiang Univesity, Heilongjiang, P. R. China 4 Automatic Contol Key Laboatoy of Heilongjiang Univesity, Heilongjiang Univesity, Heilongjiang, P. R. China E- mail: 1 climbing200310@yahoo.com.cn, 2 tcfangyuan@126.com, 3 316933876@qq.com, 4 tianguo331@126. com ABSTRACT The cold stoages ae the industial equipments fo stoing food, vegetables and othe things in low tempeatue. Chaacteistics of the cold stoage fo thei volume have a big lag and ae difficult to contol. So, taditional PID contols ae not adaptive to contol them. One of methods which can ovecome the poblems is using a fuzzy contol fo them. But it is known that the diffeential input which is needed in the fuzzy contolle is difficult to ealize and easy to poducing oscillating. A nonlinea tacking diffeentiato (TD) is substitute fo a usual diffeential input; theefoe hybid fuzzy PD (popotional and deivative) contols which include two nonlinea tacking diffeentiatos have been pesented. In this pape, the hybid fuzzy PD contol fo cold stoages with PLC is poposed to impove the system pefomances. The pape is focus on the ealizations of the contol algoithms specially. The application of the system in the fields poved the contol system elated with the pape is coect and effective. Keywods: Cold Stoage, Fuzzy PD Contol, Tacking Diffeentiato, PLC, Refigeato 1. INTRODUCTION Thee ae many cold stoages applied in the wold fo stoing food, vegetables and othe things in vey low tempeatue. One of most impotant paametes with them is the tempeatue. The efigeatos ae used to contol the tempeatue of theis. With the development and impovement of industial contol levels, the efigeatos fo cold stoages ae used wide in poductive pactice and equied bette contol chaacteistics. But fo the sake of the big volume of cold stoages and seveal point distibutions of the efigeatos and senso, the contol of tempeatue of cold stoages is vey difficult. In ode to study the poblems, some papes have been published. The fundamental factos to contol the tempeatue of a cold stoage ae to sense the infomation of the tempeatue; diffeent configuations of the cold stoages equie diffeent contolles. The embeded-systems [1] can be used fo a single cold stoage, and distibution systems [2] have been used fo lots of cold stoages. In ode to obtain the good contol popeties, some contol algoithms must be used. PID algoithm was used geneally [3], and seveal intelligent algoithms, such as simulated-human contol [4] and fuzzy contol [5] have been adopted bette. It is noted that the above contolles ae not suitable to the application because of eliability of contol systems. So the pogammable contolles (PLC) as the main contol equipment and fuzzy contol algoithms wee adopted in the pape. Additionally, PLC can ealize the othe contol functions of the system easily. As we known, taditional PID contols ae not adaptive to contol them. One of main methods which can ovecome the poblems is using a fuzzy contol. But it is known that the diffeential input which is needed in the fuzzy contolle is difficult to ealize and easy to poducing oscillating. A nonlinea tacking diffeentiato (TD) is substitute fo a usual diffeential input. Theefoe hybid fuzzy PD contols have been pesented. But, a hybid contolle can povide bette system pefomance ove a simple fuzzy contolle alone. 1811
Jounal of Theoetical and Applied Infomation Technology 28 th Febuay 2013. Vol. 48 No.3 2005-2013 JATIT & LLS. All ights eseved. ISSN: 1992-8645 www.jatit.og E-ISSN: 1817-3195 In this pape, hybid fuzzy PD contol fo cold stoages is poposed to impove the system pefomance. The pape is focus on the ealizations of the contol algoithms specially. The pape is oganized as following: section 2 deals with the design of hybid fuzzy PD contolle, section 3 pesented ealizing techniques of the contolle with PLC. Section 4 intoduces the application of contol system with PLC pesented. Section 5 gave some summay. Figue 1: The System Block OB100(Stat) FC (I/O function) FB11 FB12 OB1 (main) OB30 (inteupt function) FB10 (fuzzy contol) FB13 DB10 FB14 Figue 3: The Diagam Of Stuctue Fo Softwae Figue 4: The Diagam Of Tempeatue 2. DESIGN OF HYBRID FUZZY PD CONTROLLER The chaacteistics of the cold stoage fo thei volume ae big lag and difficult to contol. The tempeatue of the cold stoage is contolled by efigeato. Thee ae two methods of contolling efigeating media in efigeato, which ae using a suction guide valve o contolling flow speed of efigeating media to contol the speed of impelle in the centifugal efigeato diectly. In the system of ous, the fist contol method was used in ode to make use of exiting equipments. Thee is a suction guide valve in the input of the efigeato to contol efigeating capacity fo the change of efigeation load. So, the mathematic elation 1812
Jounal of Theoetical and Applied Infomation Technology 28 th Febuay 2013. Vol. 48 No.3 2005-2013 JATIT & LLS. All ights eseved. ISSN: 1992-8645 www.jatit.og E-ISSN: 1817-3195 between the tempeatue of cold stoage and the suction guide valve can be defined [4 ].Fo such contol pocess, the mathematical model of the cold stoages can be appoximately educed to fistode inetia plus time delay model expessed as below: ts G( s) = ke /( Ts + 1) (1) which T is the time constant and t is the pue delay time. Fomula (1) can be simplified appoximately as the second-ode system model: [( Ts + 1)( 1) ] G( s) = k / ts + (2) As we known, if a taditional PID contolle is applied to the fomula (2), the paametes of the contolle ae difficult to choose, and the unmodelling chaacteistics of object ae not to contol. Theefoe, the fuzzy contols wee used to impove the contol chaacteistics of system. In ode to get the easy ealization, hybid fuzzy PD contols wee designed fo the ease of diffeential contolle in physical level. The hybid fuzzy PD consists of two nonlinea tacking diffeentiatos. The nonlinea tacking diffeentiato TD ( Ⅰ ) whose input T and tacking diffeentiato output is T1 and T 2 is designed as [6-8] (4) T = T T 1 2 = R sat( T T 2 T T + 2R, δ ) 2 2 1 which a nonlinea satuation function sat( A, δ ) is defined as sgn( A), A > δ sat( A, δ ) = A, A δ, δ > 0 δ in which sgn( ) is a standad sign function. Similaly the nonlinea tacking diffeentiato TD(Ⅱ) whose input and tacking diffeentiato 1 output is and 2 can be obtained. Using the output signal of the above speed diffeentiato, the fuzzy contolle input signals ae epesented as e( t) = Ke ( T1( t) T f 1( t)), (6) (3) (5) which c( t) = Kc ( T 2( t) T f 2( t)) K and e K ae the scaling gains. c (7) So the close-loop contol stuctue which includes the hybid fuzzy PD contolle is shown in Fig 1. It consists of fuzzy PID contolle, D/A convete, electical actuato, suction guide valve, cold stoage, tempeatue senso and A/D convete. Fuzzy PID contolle is designed as follow. Supposes the two-dimensional fuzzy easoning input fuzzy language vaiable is E and C, the output fuzzy language vaiable is U. Set language vaiable E to be 6, namely {negative big (NB), negative middle (NM), negative small (NS), positive small (PS), positive middle (PM), positive big (PB)}.The fuzzy language vaiable C of change ate of linguistic value deviation is 5, namely {negative big (NB), negative small (NS), zeo (Zeo), positive small (PS), positive big (PB)}; output fuzzy language vaiable U will be 5, namely {zeo (Zeo), positive mico (PW), positive small (PS), positive middle (PM), positive big (PB)}. Rule explanation: Take the ule IF E=PB and C=Zeo then U=PB as an example, if E is (PB) and the change ate C of deviation is a zeo, in ode to obtain the quick speed of esponse, should enlage the contol quantity, namely U=PB. Altogethe 30 sentences fom the input output elations cuved suface fuzzily in Matlab, as shown in Figue 2. Figue 2: The Input Output Relations Cuved Suface 3. REALIZATIONS OF FUZZY PID CONTROLLER BY A PROGRAMMABLE CONTROLLER (PLC) Pogammable contolles (PLC) ae a most widely application in the industial contol 1813
Jounal of Theoetical and Applied Infomation Technology 28 th Febuay 2013. Vol. 48 No.3 2005-2013 JATIT & LLS. All ights eseved. ISSN: 1992-8645 www.jatit.og E-ISSN: 1817-3195 envionments. They have an advantage of high function and easy pogamming. Especially, they can easily ealize vey complex contol algoithms in one o seveal instuctions. Futhe, in the application of contolles, the eliability of them is one of the best popeties. Geneally pogammable contolles have the bette eliability than othe contolles. Duing pogamming of PLCs, usually, sequential function chat (SFC) is daw fist, and the ladde logic diagam is daw second. But, fo the complexes contol states, the integated pogammings ae mixed and not clea. So, some PLC use module pogamming, that is task of pogamming ae divided seveal pats, such as function block and data block. So, hybid fuzzy PD contolle is ealized by a pogammable contolle (PLC) of CPU 333 in seies S7-300 with type of module pogamming. Its pogam flowchat is shown in the Fig 3. Main pogam module OB1 completes the function and the function block as well as signal and data tansmission. Function module FC completes A/D sampling, caies on D/A output and othe pocessions. The ciculation inteupt oganization block fo selecting time gap is OB30 of 1s. Function module FB10 completes the entie fuzzy contol function. The backgound data block DB10 which coesponds with function module FB10, mainly saves the quantification facto and the goal tempeatue, the suvey tempeatue and so on othe paametes. FB10 is composed of fou subpocedues FB11, FB12, FB13, FB14; FB11 completes TD (1) contol function; FB12 completes TD (2) contol function; FB13 ealizes summing function; FB14 completes fuzzy contol pocessing. The function of FB14 includes caying on fuzzy pocessing, ealizing the fuzzy contol list inquiy function, completing clea pocessing. The coe technology is the fuzzy contol questionnaie inquiy pocedue. In the fuzzy contol list, fuzzy contol value U saves in block data DB10 accoding to fom ank to column and fom the left towads ight visiting way ode. The data type is intege (INT).The fist addess is DB0, is in tun DB2, DB4...DB58 (the U egula numbe is 30).If the conventional look-up table method of judgment sentence is uses, it will be able to cause the pocedue statement to be many and to be tedious, theefoe the indicato addessing table of look-up method in STEP7 will be used. In ode to simplify the design, input element in the fuzzy univese of discouse is set to incease 6, finally [0,...12].The base addess of contol quantity is 0, the displacement addess is 2 (13 Xi+Yj), by Xi, Yj may detemine the absolute addess of the contol quantity is 2 (13 Xi+Yj), obtains the U fuzzy value though the indicato vaiable which in the addess saves. 4. RESULTS OF EXPERIMENTS The system designed has been used in the field. It is the one pats of a whole industial applicable contol system with PLC. The pototype system elated with pats pesented consists of PLC, electical actuato, guide valve of efigeato and tempeatue senso. In the application, efigeating Medias uses salt wate and the salt wate inlet tempeatue is about minus five degees. In ode to confim the contol effect of the system, the ecod of the tempeatues in salt wate outlet and in the cold stoage duing 24 hou wee have made, as shown in Figue 4. It is note that the tempeatue of salt wate outlet with the efigeato is undulated minus ten degees and the tempeatue of cold stoage is undulated minus nine degees, and the magin of fluctuation was not all big, which confims the contol system designed the validity. 5. SUMMARY The hybid fuzzy PD contol with PLC fo cold stoages was pesented. Its application in the industial case was completed. The application poves the contolle pesented is effective and coect. It has the popeties of two: the algoithm pesented is bette fit to the case of the cold stoages and easy ealized. Theefoe the system esponse is speeded up, the output vibation is educed. So, it has some value of application, and makes some good effect of using fuzzy contol algoithm. It is believed that contol systems with PLC ae main tends of development fo contol systems. The futue wok is making much application of contol algoithm fo eal contol system with PLC, such as how to divide contol task to seveal modules and how to pogam ladde logic diagam fom contol algoithms. ACKNOWLEDGMENT This wok was suppoted by the Scientific and Technological Poject of Education Bueau of Heilongjiang Povince (Gant No: 12521419), P. R. China. REFRENCES: [1] Z.F. Xu, Y. Wang, Reseach on automonitoing system fo cold stoage tempeatue 1814
Jounal of Theoetical and Applied Infomation Technology 28 th Febuay 2013. Vol. 48 No.3 2005-2013 JATIT & LLS. All ights eseved. ISSN: 1992-8645 www.jatit.og E-ISSN: 1817-3195 on VB,2011 Fouth Intenational Confeence on Intelligent Computation Technology and Automation (ICICTA 2011),2011,pp.689-691. [2] A. Matí, J.C. Campelo, J. Pado, R. Os, J.J. Seano, A distibuted contol system fo citic fuits consevation and matuation based on CAN and Intenet netwoks, 2007 IEEE Intenational Symposium on Industial Electonics, (ISIE 2007), 2007,pp.1899-1904. [3] N. H. A. Hamid, M. M. Kamal and F. H Yahaya, Application of PID Contolle in Contolling Refigeato Tempeatue, 2009 5th Intenational Colloquium on Signal Pocessing & Its Applications (CSPA2009),2009,pp.378-384. [4] K. Zhang, J Zhang, G Guo, Reseach on Simulating-Human Intelligent Contol Method Fo The cold-stoage of Fuits and Vegetables, 2009 Ninth Intenational Confeence on Hybid Intelligent Systems(HIS.2009),2009,pp.413-415. [5] M.M.Rashid, A. Islam, Design and Implementation of a Fuzzy logic based contolle fo efigeating systems, 2010 Intenational Confeence on Compute and Communication Engineeing (ICCCE 2010), 2010, pp. 1-5. [6] X. Su, S.X.Yang, D. Sun, B.Y. Duan, A simple hybid fuzzy PD contolle,mechatonics, 2004, Vol.14, pp.877 890. [7] D. Wang, Hybid fuzzy vecto contol fo single phase induction moto 2010 Intenational Confeence on Computing, Contol and Industial Engineeing(CCIE2010),Vol.2, 2010,pp122-125. [8] D. Wang, Hybid Adaptive Fuzzy Vecto Contol fo Single-Phase Induction Motos, Lectue Notes in Electical Engineeing, 2012,Vol.135, p257-262. 1815