Independently rotating wheels with induction motors for high-speed trains

Transcription

1 Independently rotating wheel with induction motor for high-peed train LIANG, Bo and IWNICKI, S. D. Available from Sheffield Hallam Univerity Reearch Archive (SHURA) at: Thi document i the author depoited verion. You are advied to conult the publiher' verion if you wih to cite from it. Publihed verion LIANG, Bo and IWNICKI, S. D. (). Independently rotating wheel with induction motor for high-peed train. Control Science and Engineering,, -7. Repoitory ue policy Copyright and Moral Right for the paper on thi ite are retained by the individual author and/or other copyright owner. Uer may download and/or print one copy of any article() in SHURA to facilitate their private tudy or for noncommercial reearch. You may not engage in further ditribution of the material or ue it for any profit-making activitie or any commercial gain. Sheffield Hallam Univerity Reearch Archive

2 Control Science and Engineering Volume, Article ID 96886, 7 page doi:.55//96886 Reearch Article Independently Rotating Wheel with Induction Motor for High-Speed Train B. Liang and S. D. Iwnicki Department of Engineering and Technology, Mancheter Metropolitan Univerity, Mancheter M 5GD, UK Correpondence hould be addreed to B. Liang, b.liang@mmu.ac.uk Received September ; Accepted February Academic Editor: Ian R. Peteren Copyright B. Liang and S. D. Iwnicki. Thi i an open acce article ditributed under the Creative Common Attribution Licene, which permit unretricted ue, ditribution, and reproduction in any medium, provided the original work i properly cited. Railway vehicle with conventional wheelet often experience problem of lateral intabilitie or evere wear when running at high peed. The ue of an independently rotating wheelet (IRW) can potentially eliminate the caue of wheelet hunting and reduce wheel wear a the mechanical feedback mechanim cauing the problem i decoupled. Thi paper preent an invetigation into the deign of a novel induction motor configuration and for IRW in order to provide the tability required to atify the performance requirement for railway vehicle. A computer model of the mechanical and electrical part of the ytem wa developed. Simulation and experiment of the wheelet with active driving motor control have demontrated that a wheelet with independently driven wheel ha a good tability performance over a traditional wheelet. Controller with indirect field orientation control for dynamic control of an induction motor have hown to be uitable for thi application in both it repone and it controllability.. Introduction It i well known that railway vehicle with conventional wheel et experience problem of hunting and evere wear. The ue of independent wheel could eliminate the caue of wheelet hunting and reduce the wheel wear ince IRW can decouple the wheel. The IRW for railway vehicle ha been under eriou conideration at a theoretical and experimental level for thirty year. However, the ue of IRW would alo eliminate the guidance capability of the railway vehicle. A compromie could be achieved between the excellent tability behaviour of IRW and the guidance capability by ue of active control like yaw control, creep control or damping, and tiff control. Much effort ha been directed to the olution of thi problem. For example, independent wheel with profiled tread, independent wheel with partial coupling, or with a uperimpoition gearbox have been propoed by Kaplan [], Dukkipati [], and Jachinki [3], repectively. Alternatively eparate drive motor with precie torque control can be ued to provide guidance and curving capability (Gretzchel []). Profeor Goodall invetigated active teering and optimized control trategy for independently rotating wheel [5 7]. Although ome good computer imulation reult have been achieved, the experimental validation have not been done yet. Thi paper preent not only a ucceful dynamic control imulation for independently rotating wheel but alo validate all imulation reult by a new deigned tet rig. There are two ditinguihed apect with thi reearch. Firtly, intead of adopting very complex and impractical enor feedback ytem, a imple but effective yaw feedback control mechanim wa implemented. It could ignificantly reduce the control complexity for independently rotating wheelet. Secondly, the independent wheel were driven by induction motor which are more reliable, more powerful, cot effective and mall in ize compared to conventional dc motor. The adopt a modern vector control trategy to control the driving motor and a real-time profibu-baed control ytem ha been deigned. In Section and 3 the mechanical and electric model of the IRW are preented. Then, in Section, the implementation of the modelling by Matlab/Simulink i given. The actual deign of the tet rig and correponding-real time profibu control ytem are decribed in Section 5. Finally Section 6 and 7 dicu and give ome imulation and experimental reult.

3 Control Science and Engineering ω et ω et ω m ω m Speed Speed Motor and driving K K Motor and driving Yaw angle Figure : The overall control trategy.. Wheelet and Induction Modelling.. Wheelet Modelling. A wheelet i an important dynamic component of a railway vehicle. Generally peaking, the following equation are linear approximation for the motion of a olid axle wheelet [8]. The wheelet i taken a having the two degree of freedom of lateral tranlation motion and yaw rotation. (ẏ ) mÿ f v ψ F y mv mgϑ=, R (ẏ ) ( J w ψ f v ψ ψ l f 33 v f ψ v l ) λy =. r () The equation for IRW are very imilar to the conventional olid axle wheelet in additional to the introducing of one more tate θ, which i the differential angle of rotation of wheel. The IRW model can be given a follow: (ẏ ) mÿ f v ψ mv mgϑf y =, R ( l J w ψ f ψ v l λy l ) r r v θ l T w =, R J θ θ f ( λy l r v ψ r v θ l ) r R =. () The above et of equation decribe the behaviour of a complete wheelet... Driving Induction Motor Modelling. With the availability of fater and le expenive proceor and olidtate witche, AC induction motor drive now compare favourably to DC motor drive in the apect of power to weight ratio, acceleration performance, control flexibility, and maintenance. We adopted induction motor a driven power. Every independently rotating wheel wa deigned by an independent induction motor (Figure ). The independent induction motor at each wheel will provide the required traction and braking effort. With thi arrangement the torque at each wheel can be controlled to generate axle yaw and, therefore, teering guidance. Thi i mechanically impler than a uperimpoition gearbox. Modern vector control technique have been ued to control the induction motor. The vector control tructure are baed on flux vector determination of induction motor. There are two baic vector control tructure. The firt i flux meaurement or ening which i known a a direct field-oriented controlled drive. The econd i baed on flux modelling or lip relationhip, and thi method i known a an indirect fieldoriented controlled drive. Since the direct field-oriented controlled drive relie on integration that ha a tendency to drift, the indirect field oriented controlled drive ha been adopted for thi application. It i well known that a general d,q induction motor model with the reference axe rotating at ynchronou peed ω e canbeexpreeda[9] V e q=r I e q pφ e q ω e φ e d, V e d =R I e d pφe d ω eφ e q, =R r I e qr pφ e qr (ω e ω r )φ e dr, =R r I e dr pφe dr (ω eω r )φ e qr, φ e d =L I e d L m( I d I e dr), φ e q =L I e q L m ( I q I e qr), φ e dr =L ri e dr L m( I d I e dr), φ e qr=l r I e qr L m ( I q I e qr), T e = 3PL ( ) m φ e L dr Iq e φe qrid e. r If we let φqr e =, thi expre the field orientation concept in d, q variable. The following expreion for the electromagnetic torque, T e = (3PL m /L r )(φdr e Ie q), how that if the rotor flux linkage φdr e i not diturbed, the electromagnetic torque can be independently controlled by adjuting the tatorqcomponentiq. e Indirect field orientation i baed on the lip relationω e ω r = R r Iq/L e r Id e. If the above condition i atified, it enure that an AC motor can be controlled in the ame manner a a DC motor. Thi approach i imple and well performing for the peed and poition control even at low peed. However, the major drawback i that the orientation of the control i very enitive to the rotor reitance, which affect the robutne of the control. To overcome thi problem, the rotor reitance ha to be etimated online. The detailed configuration for driving the induction motor and inverter i hown in Figure. In order to compenate for the flux diturbance which may lightly alter the electromagnetic torque under tranient condition, a torque i introduced into the control loop. In the reearch, the wheelet i aumed to be running only on the traight track. Figure how that the yaw rotation of the wheelet i ued a (3)

4 Control Science and Engineering 3 P Lm L r Current tranform and flux etimation ω et ω m ASR ATR Speed Torque Vector tranform dq to abc tranform Current regulated PWM Induction motor T ω AΨR Field weakening Flux Figure : The vector control ytem for driving motor. an input variable to control the peed difference between the left and right wheel. The implemented whole ytem block diagram with Matlab Simulink i hown in Figure The Tet Rig Development In order to validate the imulation reult, a wheelet tet rig ha been built. The tet rig conit of /5 cale wheelet, two. kw induction motor with 7 pule/rev encoder which make accurate angular ynchronim poible, two.5 kw 6SE7 Siemen inverter, correponding data acquiition ytem, and main control computer. The whole configuration wa already hown in Figure. Itcan be een that the driving induction motor can either be independently controlled by the inverter or jointly controlled by the inverter and the main computer. Thi configuration will give the maximum flexibility and accuracy in controlling the induction motor. The trend today in automation technology are modular, reconfigure, fat, and ditributed intelligence. Therefore, the profibu wa adopted in thi application. The advantage of profibu are open, vender independent protocol, up to Mbaud, cot effective olution, and powerful diagnotic capability. The computer interface wa deigned and implemented by Labwindow oftware. The wheelet with inverter i given in Figure.. Simulation and Experimental Reult The correponding mechanical and electrical model have been implemented and imulated by Matlab Simulink oft- f = f = 8 N Table : Model parameter. f 33 = 6 N f 33 =.5Nm J w =.8Kgm J θ =.Kgm L r =. H l =.3 m L =.6 H L m =. H m=8.6kg R =.3Ω R r =.Ω r =.m P= λ=.5 ware. Table preent the correponding parameter of model ued for imulation. The fundamental idea of the IRW with yaw feedback i to ue peed difference between left and right wheel to control the yaw movement and lateral diplacement of the wheelet. The improved tability behaviour of the IRW can be determined by oberving whether the repone of the wheelet to an initial lateral diturbance. Figure 5 give one the imulation reult for a wheelet without wheel motor driving control ubjected to a 5 mm lateral pule diturbance at 3 m/ running peed. It can be een that the lateral ocillation and yaw movement of the wheelet tarted to ocillate after a 5 mm lateral diturbance wa given. Both ocillation amplitude of lateral movement and yaw movement increaed with time. The lateral and yaw ocillating frequency i about.8 Hz. Figure 6 how the correponding imulation reult under the ame condition

5 Control Science and Engineering Current regulated pwm field-oriented controlled induction motor drive Initialize and plot m3 wref??? wref in pu Clock Sum 37 Torque lambdadr e Field Weakening Tem Field Orient thetar iq e id e co rho in rho qdeabc Sa Sa Sa vag vbg vcg abcqd??? Tmech vq vd Induction machine in tationary qd iq id i Tem piqr wr/wb pidr qdabc wb wb ia ib ic Sum thetar wrm /P /P f (u) Rotor flux magnitude Gain - Gain wref??? wref in pu Sum3 37 Torque lambdadr e Field Weakening Tem Clock Field Orient thetar iq e id e co rho in rho qdeabc Sa3 Sa Sa5 vag vbg vcg abcqd??? Tmech vq vd Induction machine in tationary qd piqr pidr iq id i Tem wr/wb 3 qdabc wb wb ia ib ic Sum thetar Gain3 wheelet - Gain wrm /P /P f (u) Rotor flux magnitude Figure 3: The imulink block diagram of the whole ytem. quickly after the diturbance. Figure 7 and 8 preent another pair of imulation without and with active motor driving control but at 5 m/ running peed. It can be een that much higher ocillation for the lateral and yaw movement occurred if without active driving control after the 5 mm diturbance wa introduced. However, the ripple of lateral and yaw movement would be under control very quickly if active driving control wa applied. 5. Aement of Reult Figure : The wheelet and tet rig. but with active driving motor control. It clearly indicate that the lateral ripple and yaw ocillation will die out very A comparion between imulated and meaured peed for a wheel motor in free running i hown in Figure 9. Itcanbe een that a good agreement for it running peed and torque between imulation and experimental reult i achieved. Figure how the meaured yaw movement of the IRW with and without yaw feedback control at about 3 m/ running peed. In Figure, it i een that the IRW with yaw feedback control give much better performance than

6 Control Science and Engineering 5 Diplacement (m) 3 Input diturbance pule Diplacement (m) 3 Input diturbance pule Diplacement (m) 3 Lateral movement of the wheelet Diplacement (m) 3 Lateral movement of the wheelet Rotation degree (rad) 3 Yaw movement of the wheelet Figure 5: A wheelet without active motor control at 3 m/ running peed. Rotation degree (rad) 3 Yaw movement of the wheelet Figure 7: A wheelet without active motor control at 5 m/ running peed. Diplacement (m) 3 Input diturbance pule Diplacement (m) 3 Input diturbance pule Diplacement (m) 3 Lateral movement of the wheelet Diplacement (m) 3 Lateral movement of the wheelet Rotation degree (rad) 3 Yawmovementofthewheelet Figure 6: A wheelet with active motor control at 3 m/ running peed. Rotation degree (rad) 3 Yawmovementofthewheelet Figure 8: A wheelet with active motor control at 5 m/ running peed. IRW without yaw feedback control. The yaw movement of the IRW with yaw feedback control i barely viible while the IRW without yaw feedback control tart to how ocillation after a diturbance. The ocillation frequency i about.8 Hz which how the good agreement with the computer imulation reult preented in Figure 5. Figure and preent the wheel motor peed fluctuation under with and without yaw feedback control. The lower part of the two figure are the correponding wheel peed variation after a filter. It i indicated again that the IRW with yaw feedback provide good table capability. The /5 cale tet rig ha been teted and able to run up to

7 6 Control Science and Engineering S (rpm) andt (N m) S (rpm) andt (N m) Simulated peed and torque of the motor Meaured peed and torque of the motor Figure 9: A comparion between imulated and experimental reult for a free running wheel motor. Speed (rpm) Speed (rpm) Original motor peed without yaw feedback control Motor peed without yaw feedback control after a low pa filter Figure : The meaured wheel peed without yaw feedback control. Yaw angle (degree) Yaw angle (degree) Yaw angle without yaw feedback control Yaw angle with yaw feedback control Figure : The meaured yaw movement without and with yaw feedback control. Speed (rpm) Speed (rpm) Original motor peed with yaw feedback control Motor peed with yaw feedback control after a low pa filter Figure : The meaured wheel peed with yaw feedback control. maximum 6 mph (which equal to 3 mph in real ize) while holding good table capability. 6. Concluion In thi paper, a wheelet with independently rotating wheel with actively controlled driving motor ha been modelled, implemented, and validated by an experimental tet rig. The finding can be ummarized a follow. () Indirect field orientation control for dynamic control of motor ha been hown to be uitable for thi application in both it repone and it controllability. () The yaw feedback control alone could provide ome kind of guidance which i lot by independently rotating wheel arrangement. The imulation of a wheelet with independently rotating wheel and active driving motor control ha demontrated a better tability performance over a traditional wheelet. (3) The imulation reult have been validated, and experimental reult have hown that the propoed configuration and control methodology could lead to improved tability and good teering for a railway wheelet.

8 Control Science and Engineering 7 Nomenclature f, f, f 3, f 33 : F y : Id e,ie q,idr e,ie dr : J w : J θ : K, K: L,L r,l m : Creep coefficient Lateral force d, q current of tator and rotor Wheelet yaw inertia Wheel inertia Coefficient Self- and mutual inductance of tator and rotor Semigauge l : m: Wheelet ma P: Polepair of motor p: Differential operator R : Curve radiu R,R r : Reitance of tator and rotor r : Wheel radiu T e : Electromagnetic torque T w : Yawtorque Vd e,ve q: d, q voltage of tator v: Wheelet forward peed y: Lateral movement λ: Conicity φd e,φe q,φdr e,φe dr : d, q fluxe of tator and rotor θ: Differential angle of rotation of wheel ϑ: Trackcant ψ: Wheelet yaw movement ω e,ω r : Synchronou and rotor peed of motor. Dynamic Sytem, Meaurement and Control, Tranaction of the ASME, vol. 5, no. 3, pp , 3. [8] V. K. Grag and R. V. Dukkipati, Dynamic of Railway Vehicle Sytem, Academic Pre, London, UK, 98. [9] D.W.NovotnyandT.Lipo,A Vector Control and Dynamic of AC Drive, Clarendon Pre, Oxford, UK, 996. Reference [] A. Kaplan, T. K. Haelman, and S. A. Short, Independently rotating wheel for high peed train, SAE Paper 78, 97. [] R. V. Dukkipati, Dynamic of independently rotating wheelet-a urvey of the tate of the art, Tech. Rep. LTR-IN- 398, NRC Railway Laboratory, 978. [3] A. Jachinki, H. Chollet, S. Iwnicki, A. Wicken, and J. Von Würzen, The application of roller rig to railway vehicle dynamic, Vehicle Sytem Dynamic, vol. 3, no. 5-6, pp , 999. [] M Gretzchel and L. Boe, A new concept for integrated guidance and drive of railway running gear, in Proceeding of the t IFAC Conference on Mechatronic Sytem, vol., pp. 65 7, Darmtalt, Germany,. [5] R. Goodall and T. X. Mei, Mechatronic trategie for controlling railway wheelet with independently rotating wheel, in Proceeding of the IEEE/ASME International Conference on Advanced Intelligent Mechatronicn (AIM 5), vol., pp. 5 3, Como, Italy, 5. [6] T. X. Mei and R. M. Goodall, Robut control for independently rotating wheelet on a railway vehicle uing practical enor, IEEE Tranaction on Control Sytem Technology, vol. 9, no., pp ,. [7] T.X.MeiandR.M.Goodall, Practicaltrategieforcontrolling railway wheelet independently rotating wheel,

9 International Rotating Machinery The Scientiic World Journal Engineering Volume Senor Advance in Mechanical Engineering International Ditributed Senor Network Advance in Civil Engineering Advance in OptoElectronic Submit your manucript at Volume Robotic VLSI Deign International Chemical Engineering International Navigation and Obervation Modelling & Simulation in Engineering Advance in Acoutic and Vibration Control Science and Engineering Active and Paive Electronic Component International Antenna and Propagation Shock and Vibration Electrical and Computer Engineering