Document Sample

Speed Estimation of induction motor Drives with DSP Based Control System Solly Aryza, Zulkeflee Khalidin, Ahmed N Abdalla, Zulkarnain Lubis Faculty Electrical Engineering, University Malaysia Pahang, pekan26600, Malaysia Abstract-- Digital signal processors (DSP's) and neural important, when a slip-dependent parameter model is used network is becoming more popular in the area of ac and dc [11]. motor control, in particular, the induction motor drives. In In this case, the model coefficients of equations (1) are this paper, adaptive speed control of induction motor using difficult about impossible to be evaluated without the use of neural network inverse control scheme are proposed. The an estimation procedure [4]. controller-structure design is based on a vector control When the rotor circuit parameters are to be estimated, scheme that transforms the three phase motor currents into assuming that the stator and core parameters are known, the flux and torque generating current components. The slip-torque characteristic of the motor provides adequate experiment results show that the proposed scheme has information. However, when all the equivalent circuit excellent dynamic and static control performance. parameters are to be estimated (6) in total if constant rotor parameters are assumed and (9) if slip-dependent rotor Keywords- Speed Control, Digital Signal parameters are used, the slip-torque characteristic itself is not enough and some additional information is Processing, neural network, Induction Motor necessary[12]. Such information can be provided using, for example, the slip current characteristic of the motor along I. INTRODUCTION with the slip-torque curve. In this paper, the estimation The induction motor is a multi-variable, nonlinear, problem will be formulated in its general form, assuming strong coupled system. Its rotor parameters change very that all the equivalent circuit parameters are to be estimated prominently with the time-varying conditions. All [8]. The parameters' estimation problem can be formulated unmolded dynamic effects, such as the change of load as a least square optimization problem, the objective being torque, the existence of disturbance and magnetic saturation the minimization of the deviation between the measured make the differential geometry and inverse system method torque and current curves and the model generated curves. difficult to be applied in practice [7-10]. These curves are known as a set of discrete measurement Control method to tackle nonlinear system with uncertain points [11-13]. factor. In order to enhance the dynamic response performance of the induction motor, the differential =- + (1) geometry and the inverse system decoupling control methods are investigated. However, the decoupling and + linearization of a multivariate nonlinear system demand exact mathematics model of a controlled objects [1-2]. = - (2) Much researchers work in this field but less used dsp for media controlling. The proposed algorithm is applied to the described induction motor models and estimates the These equations are used in the estimation procedure model rotor parameters using the slip-torque motor because the motor curves and thus the measurement data characteristic. First, describes the overview of induction are obtained under nominal operating conditions. Therefore motor parameter estimation algorithm that is based on the estimation is performed under nominal conditions. numerical solution techniques [15]. The algorithm is d implemented based on the state of the art nonlinear least Vsd (t ) Rs .isd (t ) n p m (t ).sq sd (3) squares numerical solution techniques. For the estimation of dt d all the motor equivalent circuit parameters the slip-torque Vsq (t ) Rs .isq (t ) n p m (t ).sd sq (4) characteristic alone is not enough and the slip-current dt characteristic or the slip-power factor characteristic can be d additionally used, to provide extra information. Such Vrd (t ) 0 Rr .ird (t ) n p m (t ).rq rsd (5) characteristics can be most of the times obtained by the dt manufacturers. Here the characteristics are assumed to be d known as a number of discrete points [14]. Vrq (t ) 0 Rr .irq (t ) n p m (t ).rd rq (6) dt II. MOTOR PARAMETER ESTIMATION Vrd (t ) , Vrq (t ) = 0 (7) The estimation methodology makes use of data that, in general, can become available from the motor manufacturer, These equations are used in the estimation procedure or are easily measured, like the slip-torque characteristic, or because the motor curves and thus the measurement data ar the slip-current or slip-power factor characteristics [3],[5]. A Obtained under nominal operating condition. parameter estimation procedure becomes even more Therefore the estimation is performed under nominal conditions A pseudo-linear composite system can be gotten by cascading the inverse system before the original system programming of new structures on the base of recently [8]. And it was equivalent to two second-order integral debugged and tested macroinstructions. linear subsystems, so that system control of induction The programs are written in the form of macro motor which is complex multi-variable and strong coupling instructions. This method permits fast and well-arranged was transformed into two second-order integral linear programming of new structures on the base of recently subsystems control, a fully dynamic decoupling was debugged and tested Macroinstructions. achieved between flux and speed of induction motor [9-10]. sd Ls .isd (t ) Lm .ird (t ) (8) IV. ADAPTIVE SPEED CONTROL OF MOTOR INDUCTION sq Ls .isq (t ) Lm .irq (t ) (9) rd Ls .ird (t ) Lm .isd (t ) (10) The adaptive speed controller is derived from the mechanical model of system. The conversion of Eq.(2.5) rq Ls .irq (t ) Lm .isq (t ) (11) into its discrete form gives Lm . :T np (rd .isq (t ) rq .isd (t )) (12) ( J eq s Bm ). m (t ) Tem (t ) TL (t ) em Lr (16) d (13) J eq m (t ) Tem (t ) Tb (t ) TL (t ) dt Tem ( s) n p .Lm (rd .isq ( s) rq .isd ( s)) (17) Tb (t ) Bm . m (t ) (14) m ( s) ( J eq s Bm ) Lr .(J eq s Bm ) d m (t ) m (t ) (15) dt Vsd (t ) (( n p . m (t ).rq Rr L L R d rd ). m ( m 2 s ).isd (t ) isd (t ). .Ls (18) III. HARDWARE STRUCTURE OF CONTROL Lr .Ls .Lr .Ls .Lr .Ls dt SYSTEM The DSP-based control system of the three-phase Lm RL L (19) induction motor consists of three interconnected modules: V sd ( s) n p . m ( s). rq r 2m rd ( m R s .Ls .s).i sd ( s) Lr Lr Lr the eZDSP board, 161/08 DSPLINK interface and a Pulse Width Modulated (PWM) output circuit. The overall system block diagram is shown in Fig.1. n p .Lm (rd .isq ( s) rq .isd ( s)) TMF28335 DSP is used as the central processor of the m ( s) Lr .(J eqs Bm ) control system and implements the corresponding control Lm Vsd ( s) n . ( s). Rr Lm L algorithms - vector control and direct torque control. DC p m rq 2 rd ( m2 Rs .Ls .s).isd ( s) link voltage Ed, the stator currents ias, ibs, i,, ih , and the Lr Lr Lr speed w, are sampled and transmitted to the DSP through n p .Lm (rd .isq (s) rq .isd (s)) 161/08 DSPLINK interface. With the combination of these Lr Information and control methods, the required PWM gating 2 2 Lm RL n p Lm signals are generated to drive the three-phase induction J eq. .Ls .s 2 .isd (s) (( Rs ).i (s) r 2m rd ) J eqs 2 (rd .isq (s) rq .isd (s)) 2 sd motor Lr Lr Lr m (s) 1.78 (20) 3 Vsd ( s ) 0.72 x10 s 0.0157 s 3.168 2 m ( s) 2470 2 (21) Vsd ( s ) s 21 .79 s 4400 2470 K ( K P K D s ) (22) G ( s) s 2 (21 .79 2470 KK D ) s 6870 2470 KK P Fig. 1: Overall hardware architecture of control system 435 .8 ( K P K D s ) (23) G ( s) s 2 (21 .79 435 .8 K D ) s 6870 435 .8 K P The TMS320F28335 DSP used is a 60MHz, 64- bit floating point processor with two on-chip 64 bit timers, (26 an enhanced external memory interface, a two channel DMA controller and a serial port. This arrangement offers a 435 .8 K P KV lim sG ( s ) (24) versatile and powerful development tool for motor drives. s 0 21 .79 A detailed discussion of the architecture and programming 20 K P of F28 is given in [14-151. All mathematical operations of both processors are Finally, the algorithm of controller computation task can be preceded at an integer form to insure the high-speed outlined as follows: processing. The programs are written in the form of macro 1. Initialization: select a reference model. instructions. This method permits fast and well-arranged 2. Select the initial values of gain matrix K and covariance matrix P 3. Read the k" value of rotor speed obtained from encoder. 4. Compute the required control signal from the updated plant parameters along together with the information from the reference model. 5. Record the control signal u(k)given to the power module. Calculate and update the information vector. 6. Calculate and update the gain and covariance matrices. . V. Experimental System Fig3. Inverter of connection The hardware components used in this experiment consist of Digital Signal Processor (DSP) TMS320F28335 dedicated motor controller that is equipped with analog inputs, encoder inputs and PWM output channels. The PWM output channels of the DSP id fed to a power module directly linked to the induction motor. The analog inputs of the DSP are used to read the phase motor currents that are required by the inner loop current controllers. The software components consist of DSP programming tools and motor control development environment that can be used as Human Machine Interface (HMI). Estimation and control signal computation are implemented using Mat lab in the Development Tool Environment. The program is compiled on DSP compiler software and then the resulted object is downloaded to the DSP board using eZDSP subsystem. In the implementation, Fig 4: Circuit of Stator Controlling the value of measured Variables must be scaled and formatted to fixed-point representation because of the use fixed point type DSP. Scaling for the speed signal is made according to the properties of the motor encoder and DSP’s encoder input ports, where the signal is represented in bits in which one bit equivalent to 15 rpm. Under 2ms sampling time, the speed error will be 2 bits such that the speed response of the motor control system will have 2 bits of noise. VI. EXPERIMENTAL RESULT In the experiment, the speed control of the induction motor will be tested under conditions: step input set point, set-point changes as well as disturbance changes. The induction motor used in this experiment is 4poles, 60 Hz, 280V and 1.8A rating Fig5. Circuit Diagram for controlling Induction Motor Fig 2.Digital Signal Processor (DSP) TMS320F28335 Fig 10. Experimental Result of Torque Used DSP for Controlling Comparing the simulated responses (Fig.4 - Fig.6) and measured responses (fig.7 - Fig.10), it can be stated that the simulation gives relatively exact results. Some differences at the current response or Control voltage response may be caused for example by the actual properties of the real Fig 6. Estimators Adaptive in controlling sensors which are not of course ideal. Nonlinearity of the current sensors and their dynamic behavior or no concentricity of the increment sensor and motor may appear. Generally it can be stated that the correspondence between measured arid simulated responses is acceptable to Carry out sufficiently exact analysis of the drive processes VII. CONCLUSION The paper has demonstrated the implementation of this high performance direct torque control technique utilizing the floating point TI DSP, TMS320F28335. The experimental results show that an excellent torque response is achieved and agree with the theoretical and simulation results. In the end, the system is implemented on Mat lab- Simulink and DSP TMS320F28335. Fig 7. Experimental Result of current used DSP for Simulation and experiment results show that the Controlling system has good dynamic and static properties and excellent characteristic of speed tracking, and the decoupling effect between controlling Induction. For all experimental (speed responses for step set point input, increasing and decreasing set point changes as well as disturbance changes), the adaptive speed control of induction motor can give good tracking and disturbance rejection performance, where it yields zero steady state error and quick response time (most of responses have response times less than 200 ms). VIII. REFERENCE [1]. K. Kim, R. Omega, A. Cliarara, et al.. Theoretical and experimentalcomparision of two nonlinear controllers for current-fed induction motors, IEEE Trans. on Control Systems Technology, 1997, 5(3):1-11 Fig 9. Experimental Result of Speed Used DSP for [2]. R.MIarino, S.Peresada, P.Tomei. Global adaptive Controlling output feedback control of induction motors will uncertain rotor resistance, IEEE Trans on Automatic Control, 1999, 44(5):967-983 [3]. C.Kwan, F.Lewis. Robnst backstepping control of induction motors using neural networks, IEEE Trans. on neural networks,2000,11(5):1178-1197 [4]. T.K.Bonkas, T.G.Habetler.. High-performance induction motor speed control using exact feedback linearization with state and state derivatice feedback, IEEE Trans. on Power Electronics, 2004, 19(4):1022-1028 [5]. M.Rashed, F.Peter A. MacConnell, et al. Stronach. Nonlinear Adaptive State-Feedback Speed Control of a Voltage-Fed Induction Motor With Varying Parameters, IEEE Trans. on industry Applications, 2006,42(3): 723-732 [6]. A.Ba-Razzonk, A.Cheriti, G..Olivier, et al.. Field- oriented control of induction motors using neural-network decouplers, IEEE Trans. On Power Electron, 1997, 12(4):752-763 [7]. M.Mohamadian, E.Nowicki, F.Ashrafzadeh, et al. R.sachdeva,E.Evanik. A novel neural network controller and its efficient DSP implementation for Vector-controlled induction motor drives, IEEE TraNs. iND. Applicat, 2003, 39(6):1622-1629 [8]. B.K.Bose Neural Network Applications in Power Electronics and Motor Drives ——An Introduction and Perspective. Trans. On Industrial Electronics 2007, 54(1):14-33 [9]. G. Liu, Y. Sun, Y. Shen, and etc. "Dynamic Decoupling Control of Bearingless Switched Reluctance Motors Based on Neural Network Inverse System," IEEE: IEMDC, pp.1811 1815, 2005 [10]. J. Gao, and F. Li, Analysis of AC Machines and Their Systems BeiJing: Tsinghua University Press, 2005 [11] F. Lewis, A. Yesildirek, and K.Liu,”Neural Net Robot Controller with Guaranteed Tracking Performance,” IEEE Trans. on Neural Networks, vo1.50, no.3, pp.585-601, 1995 [12]. Dai Xianzhong. Wang Xin. “Neural network inverse control of current-fed induction motor”. IEEE:ISIE, p 431- 436,2008 [13]. Wu Qinghui, Lun Shuxian,and etc. “Research on neural network inverse model of induction motor drives”.IEEE: BMEI,2009 [14]. Dai Xianzhong. Liu Guohai, and etc “Neural network inverse synchronous control of two-motor variable frequency speedregulatingsystem”. IEEE: ICNSC'06, p 1070-1075, 2006 [15]. Wang XinDai Xianzhong.” The ANN inverse control of induction motor with robust flux observer based on ESO”. IEEE: ISNN, p 196- 205, 2007

DOCUMENT INFO

Shared By:

Categories:

Tags:

Stats:

views: | 233 |

posted: | 2/27/2012 |

language: | English |

pages: | 5 |

Description:
this paper not yet publishing this my own paper i sold because need money and am finish research

BUY THIS DOCUMENT NOW
PRICE: $200
100% MONEY BACK GUARANTEED

PROVIDED BY

PARTNER

How are you planning on using Docstoc?
BUSINESS
PERSONAL

By registering with docstoc.com you agree to our
privacy policy and
terms of service, and to receive content and offer notifications.

Docstoc is the premier online destination to start and grow small businesses. It hosts the best quality and widest selection of professional documents (over 20 million) and resources including expert videos, articles and productivity tools to make every small business better.

Search or Browse for any specific document or resource you need for your business. Or explore our curated resources for Starting a Business, Growing a Business or for Professional Development.

Feel free to Contact Us with any questions you might have.