Electrical Machines And Drives Present And Future

ELECTRICAL MACHINES AND DRIVES: PRESENT AND FUTU P. Vas PhD, DSc, CEng, F E E University of Aberdeen, Aberdeen, AB9 2UE, Unil ecl Kingdom Tel (44) - 1224 - 272818 E-mail P.Vas&abdn ac uk W. Drury PhD, CEng, FIEE Control Techniques Plc, St Ciles, Newtown, Powys, Wales SY 16 3AJ, United Kingdom

Abstract

-

The paper discusses the present and future of electrical machines and variable-speed drives. Although at present improved dc drives iirc emerging the market for dc drives is expected to fall and ac drives will dominate the market. Different manufilcturers have released various universal drives, and it is expected that other similar drives will ;also emerge. There will be an increase in vector-controlled drives (speed-sensorless implementations as well), permanent magnet synchronous motor drives imd different types of reluctance motor drives (switched reluctance motor, synchronous reluctance motor). Various drives, including vector and direct torque controlled drives are also discussed and compared. It is also shown that there will be a revolution in the packaging of drives and also in the application of various intelligent control techniques, (fuzzy, ineural, fuzzy-neural etc.) techniques. Drive efficiency will be :L critical issue. Most recent trends for niiaimum configuration DSP controlled drives are also discussed. Results of a survey on the future of drives ;tire also presented. This involves a large nuniber of manufacturers.

the fact that ABB lias recently marketed one specific form clf such a drive Universal drives have started lo emerge and advanccs in infelligent control techniques and packaging of drives is causing a revolution

2. T.HE PRESENT 2i.l T h e de drives field: improved drives The coweriTiona1dc motor drive continues to take a large of the variable-sptted drive inarket However, it is expected Ittat this sltare will very slowly decline Similarly to ac drivss, efforts are being made to reduce tlhe costs and to iuxease reliability and thus new and iinpro\cd dc driver; arc being introduced by various nianufacturcrs The gzneral industrial dc drive power circuits ~ J V Cclianged little since the earliest introduction of the si\.-pulse fully contd-ollcd thyristor bridgc However, this might change, if legislation to limit the liarnionic ccrntcnt of the supply current begins to have an iiupact ou the iuarkcl

2.2

1. INTRODUCTION Variable-spced drives have existed for almost ;I cciiturv Great progress has been made from the Ward-Lconhard dc drive through mercury arc rectifier controlled diivcs lo silicon-based controlled drivcs Although iniportant advarices have been made in the ac drives field, until the early 1970's. thc dyndllilc performance of dc drives was superior to those of ;IC drives [l, 21 The introduction of the so-called fieldoriented control technique (magiieti sing-flu x-oriented control) and other vector-control sclicnies (i e. rotor-fluxoriented control, stator-flux-oriented control) has brought significant changes in the field of drives. In the 1980's microprocessors have enabled the development of cost eflective digital drives and the widespread availability of' DSPs has enabled the development of a large cariety of drives with advanced features There exist basically tuo types of high perfomiance ac drives: vector drives 13, 4J and direct torque controlled drives Vcctor drives wcre introduced more tlian then twenty ycars ago and have achieved a high degree or maturity Direct torque conuolled dnves have found vanous applicatioiis 111 Uie past ten years including traction applications and at prcscnt there is an increascd intcreqt in thcsc driycs duc lo

0-7803-3 1-09-5/96/$5.00

0 1996 IEEE

67

Ac drives: vector drives, direct torque controlled drives, reluctance motor drives

Induction motors are often thc prefemed choice in v;lriablc-spccd drik e applications Although irnproved efficiency and low-speed pcrfonnauce can be obtained by using penuanent iiiagitet synchronous motors, howcvcr. urilcss thcie is a significant reduction in the cost of rarcearth pcniiauicnt iiiagitct materials, the position of thc induction motor i s secure

The PWM \ oltagc sourcc inverter, based upon IGBTs has gained a doniinnnt position in the ac drives market in the power raiigc to 200 kW due to its ease of application, good power factor arid potential to provide good dynamic perf'oIiiiancc PWM inverters are most commonly applied in inductiori motor drives. but tlicy also fonn the basis of brushlcss ac SCIVO drives Current source invcrters using thyristors contiriue to have a role for single motor drives up to Uie 3-MW range for induction niotor drives Synchrowu:j niolor clrives with CSI find application at the higher powcr range (>4 MW) and at high speed (>5000 rpiii). Cyclocorfiverlcr-l'ed ac drives for low speed/high power applications, have particular merits The static KJ-aiucr dnvc lias Poruid extensive applications where the torquc is proportional to the square of the speed, for

example, for fans, pumps and blowers. This type of load is responsible for a substantial part of the electricity consumptioii. The introduction of various vector-controlled drives has allowed the dynamic performance of ac drives to match or sometimes even to surpass that of the dc drive. By using vector control, it is possible to control separately the flux and torque producing components of the supply currents. The trend is to develop speed or position sensorless vector-controlled drives where all the idormation required is obtained by utilising the monitored voltage and/or current waveforms. Limitations on the drive dynamic are now largely set by the bandwidth with which the inverter c a i control the stator currents of the machine subjected to vector control. This in turn is determined by the supply voltage and the inductances of the machine. For further improvement of the dynamic performance, the user will have to move to permanent magnet brushless motor. Brushless motors can also be constructed with very low inertia, giving excellent speed responses. Switched reluctance [5, 61 motors have high torque-inertia ratio, enabling them to be used in servo-drives. Vector controlled drives providing high-dynamic performance, are finding increased number of industrial applications and there are many manufacturers who market these drives (CTC, Siemens, Mitsubishi, Hitachi, etc.). There are many types of implementations, direct and indirect methods, stator-flux-oriented control, rotorflux-oriented control, magnetising-flux-oriented control etc. [l- 31. There exist several speed-sensorless implementations, but future work will concentrate on further speed sensorless drives. Drives with direct torque control are finding great interest. Recently Al3B has introduced a direct torque controlled (DTC) drive, which can work even at zero speed, bbt it is expected that other manufacturers will also follow and further developments are under way for other speed sensorless implementations. Fig. 1 shows the schematic of the DTC.

y h comparator

Fig. I Schematic of a VSI-fed DTC drive

63

In a VSI DTC drive, torque control involves separate control of the stator, rotor or magnetising flux through thc selection of optimum inverter switching modes. In its

simplest Form it is required to have the flux vector in one of the six sectors of the flux hexagon. The flux arid torque errors are restricted within their respective hysteresis bands. The scheme requires flux and torque estimators. The stator flux can be obtained by integrating the terminal voltage reduced by the ohmic losses but at low frequencies- large errors can occur due to the variation of the stator resistance, integrator drift and noise. It is not necessary to monitor the terminal voltages since they can be reconstructed by using the inverter switching modes. The outputs of the flux and torque comparators are used in the inverter optimal switching table which also uses the position of the flux vector. The main features of the DTC are: direct control of flux and torque; indirect control of currents and voltages; approximately sinusoidal stator fluxes and stator currents; reduced torque oscillations; excellent torque dynamics; inverter switching frequency depends on flux and torque hysteresis bands. The main advantages of the DTC are: absence of coordinate transformations (which are required in most of the vector controlled drive implementations); absence of separate voltage modulation block (required in vector drives); absence of voltage decoupling circuits (required in voltage-source vector drives). The main disadvantages of DTC are: possible problems during starting and low speed operation and during changes in torque command; it requires flux and torque estimators (same problem exists for vector drives). Torque response times typically better than 5 ms are claimed by ABB together with high torque control linearity even down to low frequencies including zero speed. It is also claimed by ABB that the new AC drive technology rests chiefly on a new motor model with enables the computation of the motor states without using a speed or position sensor. However, the motor model used by ABB is a conventional type of model using various machine parameters (i.e. tlie stator resistance, mutual inductance. inertia etc.), but it is believed that higher robustness and precision can be obtained by utilising advanced adaptive control theories. Switched reluctance motor drlves are gaining increasing interest mainly due to the simple technology involved. However. most of the difficulty in understanding the operation and design of switched reluctance motors originates from the fact that the switched reluctance motor is doubly-salient aud highly non-linear. Switched reluctance motors have several attractive features They require only unidirectional stator winding curreuts This results in an econonlical and robust controller. The converter has totally independent circuits for each stator winding. Direct control is achieved through the stator currents. A rotor position sensor is used to switch the stator currents at the appropriate instants. The control circuit incorporates logic and data sets to provide the switching instants and also to ensure that these are optimum in some sense in relation to the application of the drive For overall performance, in particular minimum cost, the power aid control electronics and the magnetics must be designed as a

68

single entity in relation to the particular application. Future work will also be perfonned on the applications of observers to obtain real-time cstimates of the rotor position, rotor speed and electromagnetic torque and work is also in progress to improve tlieir noise and vibration characteristics since so far aspects of noise and torque ripples have not been successfully rcsolved for thc technology to be widely applied in the field of gencral drives Synchronous reluctance motors 17. 81 arc rugged, simple in structure and cheap to manufacture In particular inverter-Fed synchronous reluctance motors require simple controllers due to the absence of rotor wiindings However. the torque density, power factor and efficicncy is only lligh for high saliency ratios (Ld/Lq) arid high LdLq. However, there is a limit to tlie maximum saliency ratio, since the maximum for the direct-axis induckincc is the synchronous inductance (Ls) for cylindrical rotor case of the same stator and airgap and thc minimum for the quadrature-axis inductance is the stator leakage inductance of the machine The rotors of synchronous reluctance motors can take different forms segmental, flw barrier and &xially laminated. For seginental rotors and rotors with flux barriers tlie saliency ratio of approximately 6-7 has been achieved and with inachincs equipped with axially laminated anisotropic rotors have produced a saliency ratio of 9-12 has bcen rcportcd ‘The stator of these motors is similar to that of an induction motor. However, the magnetic structure of thc motor maybe doubly-salient similarly to switched rcluctauce motors For comparison purposes Tablc 1 shows thc main features of switched reluctance (SWRM) arid synchronous reluctance (SYRM) motors

Rectifier-inverter system: The conventional ac-ac converter for variable-speed ac motor drives is the combination of an uncontrolled rectifier and a three-phase bridge inveirter. It is likely that for the foreseeable future the rcctifier-iwerter-J~ed induction motor drive will dominate the markcl. The most siqificant dievelopment in drive technology in 11995 has been the introduction of universal drives which ;re capable of contrcdling virtually any motor in any mode of operation Control Techniques was the first in the world to producc the first universal drive in the world 1Jnidrivc: is a radically new drive, which combines sensorless vcctor open-loop, closed-loop flux vector and high-performam brushless servo technologies as a single fully enclosed product Unidrive is available in five frame sizes to 1 MW It can be directly controlled from a tontrol module panel or remotely through a serial c;ommunications interface Universal drives o€fer distinct advanlages to the user in that the control method, parameter listing mi user interface are the same for each rnetliod oP control, eliminating the need for different training progranme!; For the manufacturer of the drive, a universal drivc can lead to signifhnt rationalisation of the nianufacturmg process and hence, to lower production costs Fig 2 shows ttie Unidrive.

Table 1 SWRM Stator Rotor Stator winding Rotor winding Stator currents Control

salicnt pole salient pole concentrated

-

SYRM

smootli borc (slottcd) salicnt pole polyphase

unidirectional multi-phase balanced position-feedback position-feedback

Due to the inherent low cost and largc obtainable torqueper-volume, synchronous reluctance motors will also have a more dominant role in the future. Future will work will concentrate on vector-controlled synchronous reluctance motor development, this will also include the detailed study of the iron losses (the results are important for high speed applications), development of new topologies of rotor structures etc. There will be continued pressure for all types of drive converters to develop, and it is witli tlie ac drive that the most significant innovations are likely to occur as drive manufacturers focus on the sector of the market witli most potential growth.

Fig. 2 Unidiive (Courtesy of Control Techniques plc)

3. THE FUTURE Various future trends have also been emphasised in the previous section, bul firther detail is given below.

3.1 General, convergence of technologies The dc drivle will not dke in the near future. It is expected that there will be a growth in the dc drive market, cspccially in the field of improved dc drives, but ac drives will be dominant. Due to its ruggedness, low cost and reliability the induction motor will remain to be the preferred niacliine ]’or most industrial drives. However, higher performance arid kigher power densities will call

69

upon the permanent magnet synchronous motor drives Various reluctance motors (switched reluctance, synchronous reluctance) may have an increased role Open-loop drives with better perfonnance will be required in the future as well Further closed-loop drives will emerge

approximately 54 billion ln volume terms this corresponds to 8 million drives (about 50 % of this is classed as low-perroniiance, low-cost drives) Tables 2a and 2b show the value and volume of the various drive types, and the estimated g-owth of the world market for drives for the period 1993-2000 respcctively

New control techniques will emerge and find wide applications such as intelligent soft-computing techniques ANN-based control, fuzzy control, genetic algorithm-based control. various self-tuning techniques etc There will be a development of various observers 191 for speed-sensorless drives It is expected that intelligent sensorless vector drives will become the standard in the future These drives have many advantages besides the higli dynamic performance

Tables 2a- Value and volume of world drive market Volume (YO Value 9'0

There is the possibility for the "electroilic" motor with ai integrated motor and controller to be manufactured Reliability is a key issue of the future Totally automated manufacture is an important goal especially since greater volumes are expccted The various types of variable-speed drives are showing the signs of convergence This is mainly the result of the introduction of the IGBT The IGBT has provided the drive industry with a voltage-controlled device. with a good Safe Operating Area, which could operate off-line at a switching frequency which gives adequate bandwidth both for vector controlled drives and brushless dc motoi servo drives. Thus there is only a little practical difference between the power circuit of an inverter for an induction motor drive and that of a brushless dc servo drive The trend in brushless servo drives is therefore for off-line operation taking advantage of the relatively low cost of inverters resulting from the higher volumes in tiie ac drives market If there will be a significant reduction in the cost of rareearth permanent magnet materials, or if a type of permanent magnet is discovered, which has the perfomiance of the rare-earth pemunent magnet matenal. but with a much lower cost, then the inverter/induction motor combination could find itself under threat in many applications from the sinusoidal brushless servo niotor drive

The replacement of resolvers and expensive decoder integrated circuits with absolutc encoders is a further significant trend in servo drives when smooth rotation is critical Further advances are expected in the field of position sensing Sensorless operation, mainly by the use o f observer-based techniques (Extended Kaliiian filtcr, extended Luenberger observer etc j may also h a ~ ea significant impact It is expected that due to the large volunies involved, there will be a significant reduction in DSP prices, which will allow further DSP-controlled drives to emerge

3.2 Size of the world and European drive markets In 1993 the world market for drivcs was in the order of k7 billion [a, 3, 101 The market for the controllers was

Dc drive Servo drives Open-loop ind motor drives Low-cost drives Other

27 11 50 6 6

20 5 19 54 2

Table 2b Estimated growth of world drive market Dc drives Servo drives Open-loop ind. motor drives Low-cost drives Other

Volume %) 5 150 100 120 30

Value Dc 10 125 100 75 15

The European market for ac and dc drives comprising ac inverters. flus vector drives, dc drives and servo drives (excluding naotors) was worth $1361.9 million in 1994 and is forecast to grow by an average of 5.9%) a year to reach the value $1813.4 million by 1999. The UK market represents a sizeable portion of this total and was worth an estimated $152.9 million in 1994. Growtli in the European ac and dc drive market is forecast to peak during 1995-1996. In 1994, Siemens and ABB together with CTC have dominated the Europem drive market. It can be seen from the Table shown above that although the dc drive market is expected to slowly expand, for openloop induction motor drives and for servo-drives it is e'qected to increase significantly. There are two basic factors stimulating growth in the European drive market. Firstly, modem industrial manu€acturing technologies are placing cousiderablc demands on industrial automation to imnprovc plant flexibility. Secondly, there is a significant trend in all industry sectors towards more efficient systems.

3.3 Packaging, "electronic motors": integration of motor and inverter It is now becoming conmon practice to use an eimded or diecast heatsink as a part of the converter casing and to complete the unit with a moulded plastic unit for the walls and the front. Increased reliability is achieved by using reduced number of interconnections: rcduced coniponents and fewer assembly steps. These developments are expected to continue with the emphasis on the case of construction, enhanced reliability and good aesthetics. With a growing market, there will be a trend for increascd automation of the manufacturing process. It is expccted that in the future the inverter and motor will be integrated into a single, compact unit ("electronic

motor"). being matched to each other and optimised already at the manufacturing stage This will also result in much simpler installation, will also eliminate the need for inverter control rooms, ventilation equipment and the cabling between the motor and inverter A significant decrease in the price OF DSPs is expectcd, this may also allow an integration of the inverter and its controllers

3.4 Power semiconductors, their management IGBTs dominate the medium-power range of variablespeed drives. Since the maximal current rating of IGBT modules is around 1 kA and the voltage rating is around 2 kV, IGBTs will continue to be used instead of GTOs at the higher power levels. The MOS controlled llhyrislor (MCT) seems to be the natural successor of the IGBT However for the MCT to beconie a successful competitor l o the IGBT, it will have to be further developed A possible altemativc to the IGBT and the GTO is the ficldcontrolled thyristor (FCT) Power semiconductor module manufacturers now provide inverter bridges in a single module (complete with gate drivers and protection circuits). There is also the possibility of including the rectifier bridges and also a transistor for switching a resistor across the dc rails during regeneration. These are now known as 'intelligent' power modules

3.5 Self-tuning, fuzzy, neural, fuzzy-neuro and genetic-algorithm-based controll systems, minimum-configuration soft-computing Various types of intelligent control schemes have bcen described for dc and ac drives in the literature and it is expected that there will be a widespread use of these techniques. Fuzzy-logic based control systems i l l ] and artificial-neural-network-based control systems [ll2 - 141 seem to offer great advantages. but more conventional self-tuning schemes (e.g. pole placement controllers 1151) will also be used increasingly. For optimal response, the control loops of a drive must be tuned to take account of thc whole mechanical system (not just the motor). but the load as well. However, the drive manufacturer does not have prior information on the nature of the load, and must provide controls which allow the tuning of the systcm. Manual tuning would require a skilled operator In the future intelligent control schemcs arc expected to become standard features of drives. However minimum configuration intelligent drives will represent an important step. Such systems are under development at Aberdeen University Fuzzy logic control is receiving great interest world-wide Fuzzy logic control providcs an algorithm which can convert the linguistic control strategy based on c ~ p c r t knowledge into an automatic control strategy F u z ~ ylogic control may yield results superior to those using conventional control algorithms "he main advantages of usmg fuzzy logic controllers are' they can work well without having to construct any model of the process (or plant) and the tuning efforl can be significantly reduced

Artificial neural networks have been recently used for the identification and control of non-linear dynamic systems Thcrc are many different types of ANNs, but it is a common feature that tlhey learn from examples rather thari by utilising coanple Y algorithms required by conventional softwarc-based solutions. The leaming and adapting capability of neural networks makes them ideal foir control putposes. 'I'he ANN can be successfully applied even if the motor which is to be controlled and the load parameters are unknown ANNs can be combined with fuzzy logics and the arising fuzzy-neuro controller has ii great advantage over conventional fuzzy logic controllers the membership functions, the number of rules and the rules themselves axe lproduced by an automated process [16] The fuzzy-y-neural controller described in 1161 contains five layer:;: the input layer; a layer for fuzzy membexshilp sets, a f k z y AND layer, fuzzy OR layer, an output layer A fuuas.y-iieura1 controller enables automated desigi. In all types of intelligent control applications nlininium configuration-based systems are preferred. This leads to greater siniplicily, robustness, reliability and reduced Execution limes. Fig 3 shows experimentally obtained results for a DSP-controlled vector controlled drive containing two fuzzy logic controllers Fig. 4 shows results obtained on ihe same drive. but utilising minimum configuration h z z p controllers __---*

* , e ,

T

::I i o

11

~

r n

W I

v

( ( d d

-___

$ 1

m e I,,

~

7,m e

( 3 ,

Fig. 3 Expcrimental results for DSP-controlled ruzzy induction motor drive

3

11111'

61

JI

?I/

iuv"vvq

,+% ,.A -.

(4

- ---.w

-I IIIIC'

(\)

I

(1

Fig

11 5

1 7

2 1

DSP-controlled minimum icoifiguration fuzzy induction motor

drive

71

I

4 Experimental results for a

Genetic algorithm can also be used for motor control [17]. They are optimisation procedures inspired by the laws of a natural selection and genetics. A combination of fuzzy and genetic algorithm based techniques can also yield self-tuning controllers which automatically produce the membership functions and the rulc-base.

increasingly. The interest in photovoltaic and windpowered systems is expected to grow. Hydroelectric power generation will increase its share.

3.6 Efficiency

Bclow are extracts from reports on the future of drives provided by various well-known drive manufacturers.

The need for energy conservation is accelerating the requirement for increased levels of electric motor efficiency. It is expected that issues concerning efficiency may become increasingly important For energy saving purposes, in the future, for high-speed applications, there will be a trend to replace conventional induction motor drives employing step-up gearing with unconventional drive systems

3.7 Railway traction, electric vehicles In Europe and Japan, the activities of the traction engineers have contributed to the development of high power GTO thyristors. Traction drives using IGBT's have also been built in the past few years It would appear that in the future for high-speed traction the main competitors will be magnetic levitation systems. Recently there has been an increased interest in electrical vehiclcs However, no clear leader has emerged from the various motor-controller systems, which include chopperFed brushed or brushless dc niotors, inverter-Fed induction motors with vector control and also switched reluctance motor drives However, the key component to the success is the battery. Although at present dc motors are dominant in electric vehicle application, it is eqcctcd that in the future their role will be taken over by the induction machine. Further advances in the application of svvitclicd reluctance motors in this area are also e.\pected

3.8

Machine design, testing and maintenance, other developments

The main trends in machine design are: standardisation of structura and parts, minimisation of parts; minimisation of production costs and tooling; optiinisation of the design process. The main efforts in manufacturing are directed to automation to achieve a reduction of costs, shorter delivery times and higher quality There is a trend for computerised tests to be perfomied on thc test bed. Microprocessor-based protection and diagnostic-condition monitoring systems [ 181 will become more widely used in the future. Further developments are expected in the fields of piezoelectric actuators, ultrasonic motors. The slotless brushless motor is gaining rapid acceptance in industrial and automotive applications in the USA. The high torque hybrid stepping motor is fimding increased nuniber of applications. High-phase induction motors may be used

3.9

Future of drives manufacturers

as seen

by various

m:Most recent development has seen the introduction of a product-specific standard for power drive systems. IEC 22G/21/CDV which appears to gain approval. This standard will take precedence over all other standards prcviously used and it stipulates the required emission and irnxnunity levels of drive systems and test methods to measure the levels

Allen Bradlev/Reliance Electric: With the advent of niore advanced digital drivcs the possibility of direct reniote coinniunication with the drive itself becomes realistic This will result in a more comprehensive and immediate diagnostic evaluation. In the future it will be possible for a drive manufacturer to 'dial up' a drive via standard telephone links utilising readily available pc and modem colliigurations and also to read drive and operational diagnostic data registers and make the required adjustments remotely. Continuous remote drive interrogation and surveillance will result in more efficient setting up of drives and fewer site visits for periodic maintenance

Baldor: In addition to industrial applications. quieter. safer variable-speed drives will find increasing commercial and domestic applications with energy savings. An increasing nuniber of drives will be used in lift and escalator drives. In the commercial field further variable speed drivcs are expected for heat, ventilation control and for applications in air-conditioning systems. Further intelligent domestic drives For applications in washing machines, dishwashers etc. are expected.

Cegelec: In the future designers will need to make use of a range of techniques to limit harmonics produced by non-linear loads. These include: use of dc link chokes and ac supply reactors; connection of loads at higher voltages; incorporation of harmonic filters; increasing the pulse nunibcr of the rectifier; adoption of a common dc bus; utilisation of sinusoidal rectification systems; the deliberate introduction of distortions to the modulation pattem driving the semiconductors of the sinusoidal rectilicr to produce harmonics in anti-phase with existing disturbances etc.

Danfoss: The application of serial communications have the following advantages in drives: reduced installation time: flexibility; reduction in cabling; cost effectiveness arid flexibility. At present the most widely used fieldbus is the Profibus system delivering reliable communications

72

at speeds up to 12 Mbaud This enablcs the commissioning of a drive from a central location in 0 5 Ins, and up to 120 different equipment can be linked on the same system Larger drive manufactures at ]present can offer Profibus and WorldFP fieldbuses With advances in technology serial conmunication plays a more vital role since greater levels of control and monitoring will be required Serialconms is also ihe key to progressively smaller. easier mannfaclurabilily and thus to cheaper control systems. Serialcomms also offers significant benefits in terms of flexibility. since system changes require only changes in the software Their application also results in great benefits in the commissioning and installation procedures

Eurotherm Drives: Recently a quite revolution has taken place in the field of drive control The sophistication and adaptability of drive control has undergone enormous changes by hamcssirig digital technology, a process which is set lo continue At ]present there are highly sophisticated ac and dc drives offering the user standardised hardware that can be programmed to perform the drive functions required The next inajor step will be related to the introduction of DSPs in place of microprocessors. ']This will have two major advantages to the user: the control algorithms will beconic faster and more accurate; there will be an increase user flexibility DSP technology will place new requirements on die driver-to-user interface. Next gerxration gi-aphica? systems which are pc-based x e e ~ p c c ~ etod beconnc the norm. There will be furtlier developments of future support softwares as well in the direction of intclligent support softwares. Commissioning of drives is another area where giant steps are expected

w:

The advance of inverters has bcen inextricatebly linked with semiconductor and microprocessor technology and this trend will continue for the immediate future. At present the ac drive standard rests at tlic selftuning, sensorless torque vector level. The current generation of torque vector drives offers supdative starting torque and high dynamic perfomiance without motor feedback but falls just short in zero speed performance Flux vector drives close the gap and offer highly accurate speed control and positioning for more specialist applications, ever down to offering Iull torque at zero speed but at the inconvenience of an encoder on the motor shaft. Current products therefore offer much more than most users need for simple applications but the demands of volume manufacturing mean that the standard drive is becoming more and more complex to embrace a wider market sector with a single model Once dc drive performance levels are acllieved as an opcn loop ac drive standard, then fundamental changcs will take place Efficiency is still an area for improvement New circuit board techniques, such as double-sided surface mount technology and LSI also have made a huge inipact 011 inverter compactness and reliability The compactness of modern inverters has also led to the dcvclopment of the

inverter witliin Uie motor. This market sector is expected to e,\pand The push €or efficiency may also bring into standard production thc four-quadrant regenerative drive. taking advantage of fully integrated 'intelligent' power chips.

Mtachi. In the future there will be no need for printed manuals Many drives already have automatic tuning features Users will get advanced features whether they need tliierii or not, ;imply because the economics will dictate that every drive gets everything Several inverter manufacturers already incorporate bus systems as standard I n the future there will be a larger scale of networking and network control Fault tolerant drives and drives with rcmote diagnostic features will be more common place Nevi drives could be programmed and inslallcd electronically from remote centres The latest gencralion of intelligent IGBTs is about to go into production These will give greater protection in the current liniiting circuits offering still higher reliability. _ Mitsubishi _ _ Electric _ _ ~ A power - hybrid module is under devclopmeni that has input diodes and output transistors in the same semiconductors The near future will also see an assortment of smart input bridge topologies, including active sections designed to reduce mains disturbance to minimum They have e rcellent performance and they can pass power in the reverse direction In the future active EMC filters may replace the bulky EMC filters applied at present They are easy to design, have high performance and Lhc user can achieve space saving Future development of IG13Ts involve the integration of all control and protection electronics in the same package New dcvclopnients in the MOS controlled thyristor may change Uie cost and performance of large drives Low noise inverters are gowing in demand and the drives industry responds to this by designing low loss ICBTs which are switched at high fTequencies This considerably rcduccs audible noise, but at the cxpense of step change in the RFI (output and additional stresses on the motor windings To counter this, designers can use soft switching. but this rcquires special components or output filters At present any competent inverter on the market uses c\trcmely advanced software This is a major area for dcvclopment

Siemens. Technology in industry has taken a giant leap, but Uic level of dynamic perfomiance provided by the state-of-the-art d r i i e s is not required in many applications Probably less than 10% of all drive applications really denmnd the level of technology that is on offer __I__

Telemcran!&~ Variable speed drives are becoining more and more alike. Ftuthermore ac drives are rapidly displacing their de counterparts in all but the most specialist applicatioirj Modern inverter drives incorporate flexible microprocc 3sor systems, which means that successful features ollrered by one manufacturer can sooiibc copied lby ollicrs The important differenlidtirig

factor for the future between the various suppliers of drives will not be technology, but support The versatility of new drives will enable their use in more and more complex applications. The focus of attention will shift from the drives themselves to the customers and the applications The most advanced drives arc selfconfiguring in standard applications. however, it is their handling of the more specialised and unusual applications which is one of the factors that will ultimately separate the good supplier from the merely adequate. The successful suppliers will be customer-focused

Yaskawa. The integration requirements of modem drive systems mean communications are vital. The chosen products must offer seamless compatibility and reduced software The true aims of an integrated drive solution are. the ability to communicate a large nuniber of drives on a network; communication via a simple two-wired or three wire system to simplify installation and reduce costs, the ability to connect other pieces of factory-level equipment, easy access to a small or large iiumber of parameters in the drive line with the coniplexity of the application, a data throughout that will not degrade any of the vast variety of applications: simple software connection for ease of commissioning and ease of use, integration designed in to individual products from the outset, rather than relying on the different products of a number of different suppliers 4. CONCLUSION The paper has attempted to give a survey of the present and future developments and trends in electrical maclunes and drives. It was shown that the market share of the drives will siMvficantly increase and the cage induction machine will remain the dominant machine for industrial drives. However, there is a continued role for dc drives, especially improved dc drives. There will be an increase in sensorless high-performance drives (DTC + vector), and there will be a larger role for permanent magnet synchronous drives for applications requiring a high perfomiance or high specific output. There will also be an increased role of various reluctance motor drives for various applications. Intelligent softcomputing control techniques will become dominant. extended observers, improved self-tuning controllers will find many applications. Furtlier universal drivcs will emerge. Energy efficiency, diagnostics and condition monitoring will become more important. There will be a revolution in the packaging with integrated power modules providing a complete controller for smaller drives and maybe even the integration of the motor and the inverter

based research of Professor Vas. The authors also wish to thank Control Techniques plc for the photographs.

6 . REFERENCES [ I ] 1.' Vas, W. Drury, "Future dcvclopnients and trends in electrical machines and variable-speed derives", (In Italian), Energia Elettrica, January, 1996. 121 P. Vas, "Electrical machines and drives: A space-vector rheory approach"; Oxford University Press. 1992.

131 P. Vas, W. D m y , "Vector-controlled drives", PCIM, Numberg, pp. 213-228, 1994.

141 P. Vas, "Vector control of ac machines", Oxford lrniversity Press, 1990.

151 P. J . Lawrenson, "Switched Icluctance perspective, ICEM, pp. 12-21., 1992.

drives": A

[6] P. J. Lawrcnson, "A brief status review of switched reluctance drives". EPE Journal, 1, pp. 133-144, 1992.

17 ] A. Fratta, A. Vagatti, "A reluctance motor drive for high dynamic perfonnance", IEEE US, pp. 38-43, 1988. [SI G. Franceshini, G. P. Rosso, A. Fratta, A. Vagati, "Performance of synchronous reluctance motors in servo-drive applications", K I M , Nurnberg, pp. 15-27, 1993.

[Y] T. Du, P. Vas, A. F. Stronach, "Design and application of extended observers for joint state and parameter estimation in high-perfoniiance ac drives", E E >Pt.R, 71-78, 1995. [10] C . T. Whcatley, "Drives", 5th EPE, pp. 33-39, 1993. [1 11 Vas, I>.>Stronach, F, Neuroth. M., Du, T.: A fuzzy controlled speed-sensorless induction motor dnve with flux estimators, IEli E m , Durham, 11-13 Septeniber, 1995, pp. 3 15-319. 1121 P. Vas, J. Li, A. F. Stronach, P. Lees, "Artificial neural network-based control of electromechanical systems", Control 94, IEE, COVCIL~Y, pp. 1065-1070, 1994. [13] A. F. Stronach, 1'. Vas, "Variable-speed drives incorporating multi-loop adaptive controllers", IEE, Pt.11, pp. 41 1-420. 1995. 1141 F. Fillippetti, G. Franceschini, C. Tassoni, "Neural network aided on-line diagnostics of induction motors rotor faults", LAS, Toronto, pp. 313-323, 1993.

[15] M. Cimncione, R. Rizzo, G. Vitale, P. Vas; "Neural networks in motion control", PCIM, Numberg, pp. 131-141, 1995.

[ 161 A. F. Stronach, P. Vas, "Fuzzy-neural control of variablespeed drives", PCWI, Nuniberg, 20-22 June, pp. 117-129, 1995. [17] C . Franwsliini, A. Piuzi, C , Tassoni "A genetic algorithm approach to design flux obseiver for induction servo drives", ECON94. Bologna, pp. 2132-2136, 1994.

Extracts of a s w e y conducted on thc futurc of drives has also been presented.

[IS] P. Vas, 'Pararrieter estiniation, condition monitoring and diagnosis of eiectrical niachines", Oxford University Press. 1993.

5. ACKNOWLEDGEMENT

[19] E. Chiricozzi, F. Parasiliti, M. Tursini, D. Q. %hang, "Fuzzy self-tuning control of PMSM synchronous motor dnves", PEDES'95, Singapore, pp.749-754, 1995.

Acknowledgement is niadc to EPSRC, GEC and Control Techniques Plc for the support of the soft-computing-

74