Intenational Jounal of Applied ngineeing Reseach ISSN 0973-456 Volume 1, Numbe 19 (017) pp. 8338-8345 Soft State Induction Moto hee Phase With Fuzzy Logic Contolle Inside Delta Adhi Kusmantoo 1, heodoa Indiati W and Sigit Ristanto 3 1 Adhi kusmantoo is a lectue in the depatment of electical engineeing, Univesitas PGRI Semaang, Jl. Sidodadi imu No.4 D.Cipto Semaang, Indonesia. heodoa Indiati is a lectue in the depatment of infomation technology education, Univesitas PGRI Semaang, Jl. Sidodadi imu No.4 D.Cipto Semaang, Indonesia. 3 Sigit Ristanto is a lectue in the physics education depatment, Univesitas PGRI Semaang, Jl. Sidodadi imu No.4 D.Cipto Semaang, Indonesia. Abstact Soft state is an electonic device that can be used to contol the opeation of AC motos. he soft state consists of a thyisto component mounted anti-paallel with a thee phase AC souce. Some of the induction moto stating methods that ae mostly done befoe the use of soft state ae pimay esisto, autu tansfome, and sta delta. he method of stating pimay esisto and autotansfome poduces a lowe stating cuent when compaed to the DOL method, but the esulting toque is lowe. In addition both methods poduce powe losses on esistos o coils. Anothe stating method that is still widely used is the sta delta method. Consideing the need to contol the flow of submesible pump moto stating, it is necessay to eseach to develop soft state with geat efficiency. he pupose of this development is to poduce a soft state capable of poducing a low stating cuent but capable of poducing enough toque to stat the thee phase induction moto. In this study developed soft state using Altea FPGA with fuzzy logic contolle. he submesible pumps ae used with a powe capacity of 5.5 KW. In the sta delta method the cuent begins to each the maximum value in vey shot intevals in the sta junction fo 6 seconds and when it tavels in the delta junction thee is an incease o suge of cuent lage enough to each the full load cuent of 18 A. A consideable beat is also seen in Speed change chat. When submesible submesible pump motos wok using a soft state it appeas that thee is a gadual incease in the initial cuent ise and the moto takes geate time to each its steady state value thus minimizing moto toque pulsation. his can avoid mechanical wea, extend moto life and incease the efficiency of submesible pumps. Keywods: Soft State, Submesible Pump, Stating Sta Delta INRODUCION Soft state is an electonic device that can be used to contol the opeation of AC motos in stating and stopping motos equipped with safety o potection. he soft state consists of a thyisto component mounted anti-paallel with a thee phase AC souce, the soft state woks based on the angle setting of the thyisto o IGB o MOSF which can cause voltage changes to the moto (Suva Gupta, 014). Some of the induction moto stating methods that ae mostly done befoe the use of soft state ae pimay esisto, autu tansfome, and sta delta. In the pimay esisto method, the esisto is installed diectly fom the voltage souce befoe going into the moto coil. he cuent flowing to the coil will decease and the moto voltage dops. When the moto is otating and eaches full load speed, the esisto must be detached fom the souce of the moto cicuit. Regadless of the esisto can be done mechanically o electonically. While in the method of auto tansfome, basically the same concept with the pimay esisto method and the diffeence is only on the use of the pimay coil of the tansfome to decease the moto voltage. Moto voltage egulation of magnitude anging fom 50%, 60%, o 80% to the voltage souce (M.A. Saqib, 007). he method of stating pimay esisto and autotansfome poduces a lowe stating cuent when compaed to the DOL method, but the esulting toque is lowe. In addition both methods poduce powe losses on esistos o coils. Anothe stating method that is still widely used is the sta delta method. In this method at the stat of the moto coil connected in sta and when the moto eaches the nominal cuent with sta connection, the moto coil connection tuns into delta. he sta delta method does not cause powe loss. In the stating method using a soft state decease and voltage incease is pefomed using IGB o thyisto components (Mohamed Wasim Ansai.K, 014). he use of soft state in the industy has inceased to educe the initial flow by maintaining the toque accoding to load 8338
Intenational Jounal of Applied ngineeing Reseach ISSN 0973-456 Volume 1, Numbe 19 (017) pp. 8338-8345 equiements. Soft state based intelligent systems began to be developed to impove the pefomance of conventional state. One of the development of soft state using atificial neual netwok with back popagation algoithm. Ability to coect eos then back popagation algoithm is moe suitable fo atificial neual netwoks. he back popagation algoithm with momentum gadient paametes povides a sufficient and efficient esponse when compaed to DOL and sta delta state. he main function of the soft state is to limit the initial cuent which will potect the coils fom heat distotion especially fo squiel cage motos. his efficiency incease is done with load changes. he conventional soft state ealization simply uses the six anti-paallel thyistos component. he stato voltage setting is done by adjusting the tigge angle of the thyisto. In many cases the systems and design systems focus on the often ovelooked chaacteistics of steady state (uton Chanda Mallick, 015). Soft state in induction moto is used fo blowe, mixe, ginde, pump and othe industial usage. he basic concept of a soft state in a thee phase induction moto is by setting the incoming voltage to contol the initial cuent and the toque used. Soft state is designed using NN (neual netwok) and ANFIS (adaptive neuo fuzzy infeence system). he atificial sysaph netwok implements the feed back estimato, while ANFIS with the help of ANN estimato adjusts the tigge angle of the thyisto to contol the AC voltage unde diffeent load conditions. Implementation will give bette esults when compaed to using DSP (Syed Abdul Rahman Kashif, 007).he toque geneated by a moto at the stat is the tosion of the asul and its value can be geate o less to the otation toque at full load. s. x...(1) R R R max R. R R (. R ). max...(3). Compaison between full load toque and maximum toque is expessed by equation (4) S. R.. x max R S. L L max R. S. R S Sm S S Sm...(4) While the compaison between the toque as well as the maximum toque is expessed by equation (5) (Bos Mathew, 013). s max k R. R.. x R. s. R.. Sm...(5) max R R 1 Sm 1 s.. k R....() R he moto toque at the beginning o at the stat of the spin is expessed by the equation. I. Cos. is the oto / phase voltage at the time of otation, while I is the oto cuent / phase at the beginning. S.. R R ( S. 3 k.. Ns 3 x.. Ns R max ) S. S.. R R ( S.. R ( S. ) ) Figue 1: Chaacteistic toque against slip. Seeing the need in contolling the stating cuent of induction moto, it is vey necessay to do eseach to develop soft state with geat efficiency. he development of soft state has been stated using atificial intelligence system (fuzzy logic, neual netwok, and genetic algoithm). he pupose of this development is to poduce a soft state capable of poducing a low stating cuent but capable of poducing 8339
Intenational Jounal of Applied ngineeing Reseach ISSN 0973-456 Volume 1, Numbe 19 (017) pp. 8338-8345 enough toque to stat the thee phase induction moto (Fu Chao, 013). In this study developed soft state using Altea FPGA with fuzzy logic contolle. FUZZY LOGIC CONROLLR Fuzzy logic theoy accoding to (Lin, Chin-eng, 1996) was fist discoveed by a pofesso fom the Univesity of Califonia at Bekeley Pof. Lotfi Zadeh in 1965. Fuzzy logic is an appopiate way to map an input space into an output space. Fuzzy technology is able to solve vey complex behavio poblems and eliability is also high. hee ae seveal easons why people use fuzzy logic. o simplify complex and complex system behavio, the system in fuzzy logic uses linguistic vaiables. he linguistic vaiables in its penjaban ae expessed in natual natual language that can follow the logic o mindset of man. he use of the fuzzy set is a systematic basis fo manipulating the concept of uncetainty o obscuity. With linguistic vaiables the plant contol citeia can be made easie. he defuzzification unit does the mapping of the fuzzy infeence output to a fim value. he input of the deflection pocess is a fuzzy set obtained fom the composition of the fuzzy ules, while the esulting output is a numbe in the fuzzy set domain. he pupose of defuzzification is to poduce nonfuzzy contol action that epesents the possible distibution of fuzzy infeence contol esults, which can be expessed as follows (Mohamed Wasim Ansai, 014): y 0 = defuzzifie (y) Figue 3: he contol system stuctue applies FLC. he desied plant output (t) is compaed to the actual output c(t). o () and change of eo (C) ae input vaiables fo FLC. FLC output is a plant input o input pocess (input pocess, PI). he ule base is based on a qualitative knowledge (heuistic appoach) closed-loop system that aims to educe oveshoot, ise time, and oscillation, step-step esponses. Fo example the fuzi set fo each input / output vaiable is expessed by the tems [N, Z, P] which means Negative, Zeo, AND Positive (homas Si Widodo, 01). RSARCH MHOD Figue : he basic stuctue of the fuzzy infeence system. he fuzzification unit pefoms the fuzzification pocess of the cisp input data in the following way: 1. Mapping the fim value of the input vaiable to the appopiate discouse univese.. Convesion of the mapped data to a linguistic tem coesponding to the defined fuzzy set fo that vaiable. Fuzzification is the pocess of mapping fom the inputs of obsevations to the fuzzy set in a vaied univese of talks. In the pocess of contolling the obseved data usually in the fom of cisp and fuzzification needed to change the ange of input obsevation to the appopiate fuzzy value fo the input system vaiable. he esulting mapping data is then conveted to the appopiate linguistic fom of the defined fuzzy set fo the system input vaiables. his pocess can be expessed as follows: = fuzzifie (x 0) Data Obsevation In conducting the design of soft state fo the submesible pump moto equied data as follows: A. hee phase powe pump 5.5 KW. B. Gundfos SP 17A-10 pump bands. C. Pump capacity m3 / hou. D. Stat sta delta system.. he location of the pump in Semaang sai house housing. Figue 4: Pump location 8340
Intenational Jounal of Applied ngineeing Reseach ISSN 0973-456 Volume 1, Numbe 19 (017) pp. 8338-8345 Designing Soft State Cicuit In doing the design of soft state equied main component to suppot soft state cicuit that is: A. hee-phase contacto esys D type. B. hemal oveload elay. C. hee-phase MCB 0 A. D. MCCB thee phase 3 A.. ACS71 cuent senso. F. Optocouple MOC301. G. RIAC BCR0AM. H. Wate level senso CP35MH80. I. Altea FPGA. Modeling Matlab able 1: he input matix ouput Befoe doing the implementation in hadwae, simulation using Matlab to know the chaacteistics of soft state to be used. Soft state is designed using fuzzy logic contolle with two inputs, namely cuent and voltege. While fuzzy output is stating, the fuzzy ules use table 1. Both the input and the output use the seven membeship functions of the same ange. Figue 5: Design soft state inside delta he poposed design in this study used a soft state inside delta, with the aim of eplacing the conventional sta-delta state that often induces a pounding of the induction moto. In the delta connection is used schneide contacto with type esys D. Wate level sensos ae used to povide infomation to the contolle of wate availability, stat time and stop time. Cuent senso seves to povide infomation on the magnitude of induction moto stating cuent. he tigge cicuit seves to tigge RIAC in adjusting the moto voltage. Figue 7: Input cuent Figue 8: Input voltage Figue 6: Block diagam of a single phase soft state 8341
Intenational Jounal of Applied ngineeing Reseach ISSN 0973-456 Volume 1, Numbe 19 (017) pp. 8338-8345 RSULS AND DISCUSSION Figue 9: Output stating Figue 11: Soft state algoithm using fuzzy logic Figue 10: Fuzzy logic ules o ecove the data fimly used deffuzification with Methode Centoid (Composite Moment). In this method the data is fimly obtained by taking the cente point of the fuzzy egion. z COA Z ( z) zdz z A ( z) dz A he pocess of stopping the induction moto is done subtly by educing the voltage of the thee-phase electic souce to zeo o at the initial voltage. his pocess occus when the wate in the containe is nea full, the wate level senso will detect the wate level in the containe so that the submesible pump induction moto must stop woking. In addition the cuent sensos also povide infomation to the FPGA duing the pocess of induction moto cuent decline. he method of stopping the opeation of the submesible pump moto with soft stop baking povides an efficient way with no need fo additional cicuits in the design of this soft state. Because baking by educing the voltage of the voltage souce is done slowly so as not to cause a loss of stoke on the induction moto and load. 834
Intenational Jounal of Applied ngineeing Reseach ISSN 0973-456 Volume 1, Numbe 19 (017) pp. 8338-8345 cuent to each Full load cuent 18 A. A substantial beat is also seen in the speed change gaph in figue 14. When submesible pump motos wok using a soft state it appeas that thee is a gadual incease in the initial cuent ise in figue 15 and the moto takes geate time to each Its steady state value theeby minimizing moto toque pulsation. his can avoid mechanical wea, extend moto life and incease the efficiency of submesible pumps. Figue 14: Moto speed submesible pump with sta delta Figue 1: Algoithm stoping soft state. Figue 15: Stat soft state Figue 13: Stat sta delta cuent Figue 13 shows that fo cetain time intevals submesible pump motos while woking using the sta delta system of cuent begin to each maximum values in vey shot intevals in sta connections fo a time of 6 s and when switching in the delta connection thee is a significant incease o suge of Figue 16: Moto speed submesible pump with soft state 8343
Intenational Jounal of Applied ngineeing Reseach ISSN 0973-456 Volume 1, Numbe 19 (017) pp. 8338-8345 CONCLUSION he sta delta method povides a lage initial cuent, while the soft state method povides smoothe moto otation acceleation. Soft state motos ae able to make a steady ise in cuent when the stating moto. By setting the angle of the thyisto tigge on the soft state compaed to the sta delta method, it appeas that the angle tigge setting can impove the stating cuent and the speed of the submesible pump moto. In addition, the angle tigge settings thyisto can poduce toque and moto speed accoding to the load used. he pump moto esponse gaph geneates the stating cuent within 14 seconds with the oveshoot below 1%, the moto toque esponse within 0-0. seconds changes between positive and negative and due to initial stating value with a small tigge angle. fficiency of electic enegy consumption in the submesible pump moto can be achieved by adjusting moto speed though contol of soft state. ACKNOWLDGMN he authos wish to thank the LPPM Univesitas PGRI Semaang (UPGRIS), especially Ristekdikti which has povided budgetay costs of eseach. RFRNCS [1] A.R. Kashif and M.A Saqib, Soft stating of induction moto using Adaptive Neuo Fuzzy Infeence System, Poceedings of theintenational Confeence on lectical ngineeing, Lahoe, 007. [] Adel Gastli and M.M. Ahmed, 005, ANN-based soft stating of voltagecontolled fed IM dive system, I ans. negy Conves., Vol. 0,No. 3, pp. 497-503, Sept. 005. [3] Allen Boldt, Using SCR switching cicuit in Application to moto equiement speed egulation, toque capabilities and stopping methods, www.thyistosc.co/news/z. Retieved Jan. 10, 015. [4] B.Sheeba Rani, S.Jayanthi and C.Yuvaaj, lectonic Soft State fo Induction Moto, Intenational Jounal of Advances in ngineeing, 1(3), 135 140, 015. [5] Bos Mathew Jos. K.S.Haeesh, hyistoised State fo Induction Machine Dive, ansactions on ngineeing and Sciences ISSN: 347-1964, 013. [6] C. hanga Raj, 009, negy fficient Contol of hee-phase Induction Moto A Review, Intenational Jounal of Compute and lectical ngineeing, Vol. 1, No. 1, Apil 009. [7].Rampasath, P.Manojkuma, 015, Modelling and Analysis of Induction Moto using LabVIW, Intenational Jounal of Powe lectonics and Dive System (IJPDS) Vol. 5, No. 3, Febuay 015, pp. 344~354. [8] Fu Chao, Design and Simulation of hee-phase AC Moto Soft-stat, Intelligent System Design andngineeing Applications (ISDA), pp: 554-557, 013. [9] K.Sundaeswaan, PSR Nayak, Paticle Swam Optimization Based Feedback Contolle Design fo Induction Moto Soft-Stating, Austalian Jounal of lectical & lectonics ngineeing, vol.11, No.1, Mach 014 pp.55-63, 014. [10] Koes Indakoesoema, Influence On Moto Cicuit Pump Pime RSG-Gas, Poceedings Semina Reseach and Management of Nuclea Devices, Acceleato echnology Cente and Pocess Mateials, Yogyakata, 011. [11] Lin, Chin-eng, dan Lee, Geoge. Neual Fuzzy Systems, London: Pentice-Hall, 1996. [1] Li Kai. Chen ing lin. ang Qiang, Dynamic Identification and Contol of IM Soft-StatUsing ANN, Depatment of Contol Science and ngineeing Habin Institute of echnology, Habin, PRC, 978-1-444-1706-3, 008. [13] M.A. Saqib, A.R. Kashif and ehzeeb-ul-hassan, Soft state of induction moto using neual netwok based feed back estimato AUPC 007, Peth, Westen Austalia, Decembe 007. [14] Mohanasundaam K, Sugavanam KR, Senthilkuma R. A PSO, Algoithm based PI contolle design fo soft stating of induction moto, Intenational Jounal of Applied ngineeing Reseach; 9(4):5535 4, 014. [15] Mohamed Wasim Ansai.K. Mohanasundaam.K. Mohan.C, Fuzzy Logic based Soft Stating of hee Phase Induction Moto, Intenational Jounal of Innovative Reseach in Science, ngineeing and echnology, Volume 3, Special Issue, Apil 014. [16] Osowski, Stanislaw; dan Linh, an Hoai, Neuo-fuzzy SK Netwok fo Appoximation, 004. [17] Ramamooty, M. and Aunachalam, M., 013, A solid state contolle fo slip ing induction motos,i Annual meeting on industial applications,1 octobe, 013. [18] Shweta.V. Muttagi D. M. S. Aspalli, Reduction in oque Pulsations duing Stating, IJSRD - Intenational Jounal fo Scientific Reseach & Development Vol. 3, Issue 0, 015. [19] Syed Abdul Rahman Kashif and Muhammad Asgha Saqib, Soft Stating of Induction Motos using Neuo Fuzzy and Soft Computing, I, Second Intenational 8344
Intenational Jounal of Applied ngineeing Reseach ISSN 0973-456 Volume 1, Numbe 19 (017) pp. 8338-8345 Confeence on lectical ngineeing 5-6 Mach 008. [0] Sneha M Mukae, IGB based Induction Moto Soft State, IOSR Jounal of lectonics and Communication ngineeing (IOSR-JC) ISSN: 78-834, 013. [1] Suva Gupta, Fuzzy Logic Based Soft Stating Of Induction Moto With Cuent Contol. hesis, Depatment Of lectical ngineeing National Institute Of echnology, Roukela, 014. [] Sundaeswaan, K. and Nayak, P.S.R, Design of Feed Back Contolle fo Soft-stating Induction Moto Dive System Using Genetic Algoithm, Int. J. Industial lectonics and Dives (IJID), Inde Science Publishe, Vol. 1, No., Mach 014, pp.111 10, 014. [3] Si Kusumadewi, Neuo-Fuzzy Integasi System Fuzzy dan Jaingan Syaaf, Gaha Ilmu, Yogyakata, 006. [4] Slobodan Michevski, negy fficiency in lectic Dives, lectonics, Vol. 16, No. 1, June 01. [5] Siyu Leng, R. Ul Haque, N. Peea, A. Knight, J. Salmon, Soft Stat of Induction Motos using Floating Capacito H-bidge Convetes, 7th I intenational confeence on Powe lectonics, Machines and Dives, Mancheste, UK, 014. [6] Sun Zhiping, Moto soft-stat technology Review, Jounal of Jilin Institute of Chemical echnology, pp.56-60, 013. [7] Sachin S. Mesty, Fault oleant Opeations in Soft State fo Induction Motos, Poceedings of SixthIRAJ Intenational Confeence, Octobe 013. [8] homas Si Widodo, Fuzzy Neuo System Fo Infomation Pocessing. Modeling and Contol, Gaha Ilmu, Yogyakata - Indonesia, 01. [9] uton Chanda Mallic Shaith Dha. Jubae Khan, Atificial Neual Netwok based Soft-State fo Induction Moto, Poceedings of Intenational Confeence on lectical Infomation and Communication echnology (IC), 015. [30] Zheng Qiang Wang, Qi Chao Zhang, Reseach and Design of AC Moto Fuzzy Soft-Stat Contolle,Sensos & ansduces, Vol. 175, Issue 7, July 014, pp. -6, 014. 8345