INDUCTIVE power transfer (IPT) systems have found application

3370 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 54, NO. 6, DECEMBER 2007 A Three-Phase Inductive Power Transfer System for Roadway-Powered Vehicles GrantA.Covic,Senior Member, IEEE, John T. Boys, Michael L. G. Kissin, Student Member, IEEE, and Howard G. Lu Abstract The development of a new three-phase bipolar inductive power transfer system that provides power across the entire width of a roadway surface for automatic guided vehicles and people mover systems is described. A prototype system was constructed to verify the feasibility of the design for a number of moving loads (toy cars). Here, 40 A/phase is supplied at 38.4 khz to a 13-m-long test track. Flat pickups are used on the underside of each vehicle to couple power from the track to the vehicle. Finite element modeling software was used to design the geometrical position of the track cables and to predict the power output. This design resulted in a considerably wider power delivery zone than possible using a single-phase track layout and has been experimentally verified. Mutual coupling effects between the various track phases require additional compensation to be added to ensure balanced three-phase currents. Index Terms Electromagnetic coupling, electromagnetic induction, energy conversion, road vehicle electric propulsion. I. INTRODUCTION INDUCTIVE power transfer (IPT) systems have found application where energy transfer to moving vehicles without mechanical contact is required [1], [2]. Such systems include clean rooms, monorail transportation, automatic guided vehicles (AGVs), and battery charging. In many of these, the vehicle is designed such that power is transferred continuously while it moves along a track or rail [3], [4], requiring good horizontal and vertical alignments between the power pickup and the track to ensure continuous power delivery. In monorail applications, this requirement is easily met as the magnetic power pickup is placed on a bogie whose movement relative to the track is naturally constrained; consequently, there exist numerous well-developed commercial systems [1], [2]. Roadway vehicle applications using IPT have been proposed for more than two decades [5] [7], but to date there has been only limited commercial development due to the difficulties in transferring sufficient power to a moving vehicle without imposing constraints on vehicle movement. Commercial AGVs require automatic steering control to ensure the power pickup is properly aligned with the track [8], and battery charging systems in people movers have been successfully employed but either require user plug-in or position alignment systems to ensure adequate charging [9], [10]. Roadway vehicles driven by human operators cannot meet the tolerance demanded by present sys- Manuscript received November 29, 2006; revised May 23, 2007. The authors are with the Department of Electrical and Computer Engineering, The University of Auckland, Auckland 1142, New Zealand (e-mail: ga.covic@auckland.ac.nz). Digital Object Identifier 10.1109/TIE.2007.904025 Fig. 1. Typical IPT system. tems, and consequently, system performance is compromised. Either the vehicle pickups have to be oversized or multiple pickups must be placed underneath the vehicle to compensate for the inevitable misalignments of the pickup relative to the track [11]. Alternative suggestions to overcome this problem include sequential excitation of short track segments requiring multiple switched primary coils and vehicle sensing [12], [13]. Normally, on-board batteries are also required to facilitate starting, manage power fluctuations, and enable unconstrained movement [5] [8], [12], [14]. In this paper, a multiphase IPT system that removes the aforementioned restrictions while minimizing each vehicle s on-board power pickup weight is proposed. In AGVs and people mover systems, such a system will allow vehicles to pass, which at present is impossible without on-board energy storage. Other advantages include the following: 1) providing short power boosts to battery-powered electric vehicles where increased power demand is necessary (such as climbing particularly steep slopes) and 2) continuous power transfer without onboard batteries or battery charging along roadways without the need for complicated pickup structures or driving restrictions. This paper begins by discussing the basic principles of IPT. Following this, a practical three-phase IPT resonant supply is presented along with methods of controlling and balancing the current in each phase. Finite element modeling (FEM) is then used to help determine the appropriate track spacing for a prototype system, and results are compared with actual measurements on the track. II. PRINCIPLES OF IPT A typical IPT system is shown in Fig. 1. It is composed of two distinctly different electromagnetic systems as follows: 1) a power supply takes (three-phase) power from a mains supply and energizes an extended primary loop or track, and 0278-0046/$25.00 2007 IEEE

COVIC et al.: THREE-PHASE INDUCTIVE POWER TRANSFER SYSTEM FOR ROADWAY-POWERED VEHICLES 3371 2) pickup coils or secondaries along that track have power coupled into them, which is then processed and used to drive motors, lights, or other loads as required. As shown therefore, an IPT system is a magnetically coupled system, but one where the coupling coefficient is much lower than other familiar magnetically coupled systems, such as induction motors or transformers. The power supply used drives a constant current in the primary track inductance. Low coupling factors between the track and pickup controllers within IPT systems make power transfer difficult and the cost per kilowatt high. To improve the power transfer capability, it is therefore common practice to increase the operating frequency considerably above usual mains frequencies and to use resonance with either or both the primary loop and the secondary pickup. Either fixed or variable frequency power supplies can be chosen; however, fixed frequency supplies have the advantage that the pickup design is simplified given that the track frequency will not shift under all loading conditions [1], [15], [16]. In both IPT and induction heating applications, LCL resonant topologies are commonly used as the fixed-frequency current-sourced supply [17] [19], using a full bridge with series output inductance followed by the parallel tuned track. The tuning capacitor C is chosen to resonate with the track inductance L at the operating frequency. If the supply inductance is also chosen to match the track inductance, then the network transforms a constant voltage to a constant track current and converts load at unity power factor while also functioning as a bandpass filter, reducing the generated noise. The addition of resonance complicates the IPT system. As stated, on the primary side, a resonant capacitor can provide an excitation current to the track in the same way as a transformer has a magnetizing current so that the power switches only need to supply the real power required by the load. At the pickup(s), resonance boosts the power output. The open-circuit voltage induced in the pickup coil as a result of current I 1 in the track is given by V oc = jωmi 1 (1) where M is the mutual inductance between the track and the pickup coil (self-inductance L 2 ). The short-circuit current of this pickup coil is V oc limited by the impedance of the secondary winding, i.e., I sc = MI 1. (2) L 2 A measure of the uncompensated voltampere of the pickup is given as follows: S u = V oc I sc. (3) If the secondary coil is tuned for resonance at the track frequency ω using an appropriate capacitor (C 2, either in series or in parallel), the voltages and currents in both C 2 and L 2 increase by a nominal factor of Q 2 and the power transferred from the track to the pickup coil is then expressed as [1] P = S u Q 2 = ωi1 2 M 2 Q 2. (4) L 2 Fig. 2. Common pickup shapes. (a) E. (b) Flat-E. (c) Simple flat. The output of the pickup is normally a regulated dc voltage using a switched converter that causes the operating Q 2 of the circuit to vary with load, within some maximum value (typically 10) because of limitations in control and device ratings [1], [15]. Thus, for a given circuit Q 2, a higher output power is normally possible by increasing the primary supply frequency and track current or by improving the magnetic coupling to the secondary pickup. In practice, however, nominal track frequencies of 10 20 khz are used for systems above 10 kw, as limited by available power switches; 40-kHz systems are typical in the 100 W 10 kw region and so forth. In monorail IPT systems, a common pickup configuration is the E-pickup as illustrated in Fig. 2(a) [1], [20]. Because the IPT track wires are placed in close proximity to each other, the pickup is designed such that the field contributions from both wires couple in a constructive manner into the secondary power pickup, thereby improving the power transfer capability of the system. In vehicle applications, the track wires are normally buried in the ground, and the permeable material within the pickup cannot extend into the track. In such cases, either flat-e or simple flat pickups are used as shown in Fig. 2(b) and (c), respectively. When the vertical and horizontal power profiles of these pickups are considered, taking into account potential horizontal misalignment from the track center as shown in Fig. 3, there are marked differences. In the case of a simple flat pickup, the ferrite and coil are configured to capture the horizontal flux component of the field above the road surface. The field contribution is at a maximum when the pickup is centered above any one wire and is at a minimum in the center point between the two wires [Fig. 3(b)]. For a given track width, the tolerance to horizontal misalignment is maximized by separating the track wires and using two pickups. A flat-e pickup is configured to capture the vertical component of the flux, which is a maximum between the two wires. The power capability of the flat-e is best when correctly aligned with the track; however, this profile drops considerably with small horizontal misalignments [Fig. 3(a)]. Because simple flat pickups are able to perform considerably better with misalignment, they are therefore preferred where horizontal vehicle movement cannot be easily constrained [5], [6], [10]. If the vehicle movement in a given system is sufficiently constrained, a pickup with tighter coupling to the track can be used. This allows the use of very tightly coupled pickups, such as coaxial pickups [7], for applications where the vehicle moves along a track or is stationary during the power transfer, such as a battery charging system [9], [10]. The disadvantage of these

3372 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 54, NO. 6, DECEMBER 2007 Fig. 3. Comparison of flat pickup power profiles, with 60-mm track spacing. (a) Flat-E. (b) Simple flat. pickups is of course that they inherently disallow movement or misalignment of the vehicle while power is being transferred. III. CONCEPTUAL THREE-PHASE SYSTEM In the proposed system, the objective is to produce a track layout that provides increased power delivery with greater horizontal tolerance for a given pickup topology. A polyphase IPT system that can achieve this goal is defined as being any system in which the currents in the track cables are electrically separated by angles other than an integer multiple of 180 such that a rotational magnetic field is produced around the track. Singlephase distributed-winding (where several wires are spread out to create any particular phase), sequential concentric-coil, and meander-coil type tracks [12], [13] are naturally excluded, in that these track topologies have currents that remain separated by 180 and produce a stationary field across the track. A given IPT track system can be labeled as either unipolar or bipolar, depending on whether the pickup is exposed to only the forward currents or both the forward and reverse currents of each phase. For example, a three-phase system can utilize either three track cables (unipolar) with a closeddelta output transformer [Fig. 4(a)] or six track cables (bipolar) with an open-delta output transformer [Fig. 4(b)]. While both systems have been studied, this paper considers the bipolar system in particular with the track phases laid out as shown in Fig. 4(c). An N-phase IPT system is spaced in a manner similar to a linear induction motor such that each adjacent track cable carries a current separated by 360 /N. This produces a timevarying field that sweeps across the track horizontally, allowing a simple flat pickup to capture the horizontal component of this field across the entire width of the track. In the proposed three- Fig. 4. (a) Unipolar 3Φ system with closed-delta output transformer. (b) Bipolar 3Φ system with open-delta output transformer. (c) 3Φ bipolar track layout as driven by a positive phase sequence inverter with ferrite pickup. phase system, the phase displacement between the currents in adjacent track wires is either 120 or 60 depending on whether a unipolar or bipolar arrangement is used so that the horizontal field component does not directly cancel between each wire. This can be compared to the horizontal component captured from a simple flat pickup in a traditional single-phase system [as shown in Fig. 3(b)] that contains statically positioned nulls in the power transfer profile. The physical separation of the wires and the presence of ferrite in the pickup [Fig. 4(c)] distort the field vectors so that a simple field analysis will not accurately predict the received power. Consequently, FEM modeling was undertaken to help determine the output power for a given ferrite block. This will be discussed in later sections. IV. PROTOTYPE THREE-PHASE POWER SUPPLY A three-phase inverter power supply was used as shown in Fig. 5. This system is composed of a three-phase full-bridge converter, a three-phase isolating transformer, which is part of the LCL network in each phase, a control circuit, and opencircuit protection circuitry (not shown). Here, the three-phase transformers provide galvanic isolation to the system. Using an open-delta arrangement on the transformer secondary enables all phases and their return paths to be available at the output. In practice, the isolation transformer is constructed using three separate single-phase transformers placed directly between the bridge converter and the track. These transformers have some leakage inductance and are designed to form the first L of the LCL network in each phase. Since most of the reactive current circulates in the last CL of this network, only real power goes through the transformers and the voltampere rating

COVIC et al.: THREE-PHASE INDUCTIVE POWER TRANSFER SYSTEM FOR ROADWAY-POWERED VEHICLES 3373 Fig. 5. Resonant inverter and track layout. of each transformer is small. In order to minimize conducted noise generated by the inverter, the primary and secondary coils are separated with a 2-mm gap and individually wound. This reduces the interwinding capacitance but produces a large value of leakage inductance in each transformer. If there is a volt-second imbalance due to unequal conduction voltage drop or unequal switching times in the power switches, then a dc voltage component can be present at the inverter output that will saturate the transformer cores. Capacitors C 1 to C 3 are used to prevent such saturation but are also designed to compensate for variations in the transformer leakage inductances. Capacitors C 4 to C 9 are designed to tune the various track inductances L, which should all be made identical to ensure balanced phase currents. The value of these capacitors depends on the transformer turns ratio. A. Novel Voltage Controller In normal operation, a bridge inverter is operated using a standard six-step pulse enabling a fundamental line-to-line voltage given in terms of the dc voltage V d by 6 V LL = π V d. (5) A problem with such a topology is that the inverter cannot control the magnitude of the output ac voltages, requiring the dc input voltage to be controlled in order to control the output voltage magnitude and, consequently, the magnitude of the currents in each phase. A pulsewidth modulation (PWM) technique is presented here to vary the output voltage magnitude removing the need to adjust the input dc voltage. Here, the voltage output of each phase is essentially a square wave except for a notch and a pulse to control the ratio of the third harmonic content in the output. The phases A and B output voltages V AN and V BN and the line-to-line voltage between these phases V A B are plotted in Fig. 6(a). A notch is introduced in the first half-cycle at 60, and a pulse is introduced in the other half-cycle at 240 in each phase waveform. The width of each notch and pulse are made identical and can be adjusted from 0 to 60 using a microprocessor. These variations are indicated by arrows in Fig. 6(a), along with the effect on the line-to-line voltage Fig. 6. (a) Variable output voltage control. (b) Measured PWM gate waveforms. V A B. If the width of these notches and the pulses is set to 0, the output voltage will be identical to that created using the standard six-step waveforms having a maximum fundamental value given by (5). If their width is adjusted to 60, then the output line-to-line voltages will have in-phase common-mode third harmonics, and since the delta arrangement of the isolation transformer presents infinite impedance to triplen harmonics, no current will flow in the track. In practice, the track currents can be measured and adjusted by varying the width of these additional pulses and notches giving full current control. Fig. 6(b) shows the measured gate signals to the power switches Q 1, Q 2, and Q 3 of Fig. 5 that produce the required output phase voltages. If the width of the notch is defined as θ and the voltages are normalized to have a magnitude of 1, the harmonic content of the line-to-neutral output voltages V AN and V BN and the lineto-line output voltage V A B can be obtained by a Fourier series expansion [defined in (6)] of the waveforms in Fig. 6(a), i.e., V (x) =a 0 + n=1 [ a n cos ( 2πnx T ) + b n sin ( 2πnx T )]. The coefficients of the harmonics in the line-to-neutral voltage are given by a 0 =0.5 a n = 1 ( ( nπ sin nπ 3 b 0 = 1 ( ( nπ cos nπ 3 ) ( nπ ) ( ) 4nπ sin 3 + nθ +sin 3 + nθ )) (6) ( 4nπ sin 3 ) ( nπ ) ( ) 4nπ cos 3 + nθ +cos 3 + nθ ( ) ) 4nπ cos +cos(nπ) 1. (7) 3

3374 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 54, NO. 6, DECEMBER 2007 The first five nonzero harmonics for the line-to-neutral and line-to-line voltages are shown graphically in Fig. 7(a) and (b), respectively. Note that in both graphs, the magnitude of the fundamental component decreases approximately linearly from its initial value to zero as the pulsewidth increases from 0 to 60. It is this fundamental component that controls the track current. From Fig. 7(a), it can be seen that as the pulsewidth increases, the fundamental component is replaced entirely by triplen harmonics, which are blocked by the delta-connected output transformer. The remaining harmonic components, which are not blocked by the output transformer, are attenuated by the filtering properties of the LCL network [21] so that the track current is essentially purely sinusoidal. In an IPT system, all of the track wires are fully insulated (in some cases buried in concrete), and consequently, accidental short or open circuit faults, while possible, are extremely unlikely. If either condition were to occur, however, the current in the bridge would slowly increase with a rate of rise limited by the relatively large inductances at the output of the bridge (corresponding to the large leakage inductances of the transformers as aforementioned). If the currents are not interrupted, they will eventually damage the power switches. The required protection circuit as shown in Fig. 5 is extremely simple. The slow rate of rise is easy to detect using three current transformers in the output phases of the inverter. Over current in any one phase causes a trigger signal that shuts down the inverter if a fault occurs. Upon receiving the signal, the microcontroller attempts to restart the inverter after a short delay. If the fault persists, the inverter will shut down again. In practice, the three-phase track currents within the constructed prototype were found not to be exactly balanced, as shown in Fig. 8(a). This is largely due to an imbalance in the mutual coupling between the relative phases. During operation, each phase couples voltage into the adjacent phases, but the asymmetries in the cable positions mean that these coupling factors vary. The effect of the magnetic coupling is to change the effective inductance (apparent length) of each phase, and this effect is more noticeable in the two outer phases. In order to balance all phases in the system, additional inductance can be added into the A and B phases, while phase C may require a series capacitor to reduce its overall inductance to match. The resulting balanced currents are shown in Fig. 8(b). Fig. 7. Harmonic content of the output voltage with changing pulsewidth. (a) Line-to-neutral voltage. (b) Line-to-line voltage. The line-to-line voltage is an even function so that all b n coefficients are zero. The a n coefficients can be shown to be given by a 0 =0 a n = 2 ( ( n ) sin nπ 6 (2π 3θ) sin ( nθ 2 ) ( ( sin n π θ )) 2 +sin( n 6 (4π 3θ) ) ). (8) V. P ICKUP AND TRACK DESIGN For the purpose of constructing a prototype pickup, one or more ferrite blocks with similar characteristics to 3C85 were used as they were readily available. The dimensions of each ferrite block are given as follows: length =33mm, width = 50 mm, and thickness = 3 mm. The initial pickup winding comprised 20 turns of 0.7-mm enameled wire as illustrated in Fig. 9(a). Alternative pickup configurations used two ferrite blocks with a single identical core [Fig. 9(b)] or dual 10-turn series-connected windings [Fig. 9(c)]. The prototype vehicles used here are radio-controlled TAMIYA 1 : 10 scale quick drive sports car models. In order to supply the car with the equivalent power as the usual battery, the pickup should be capable of delivering 30 W. The dimensions of the car are given as follows: length = 440 mm, width = 180 mm, and height = 140 mm, with 10-mm ground clearance. The track width is 2.5 times that of the car. Notably, the position of the pickup above the cables, the magnitude of the track current, and the cable spacing are three major factors that affect the track design. As the height of the car is fixed, then the position of the ferrite above the cables is also fixed at 6 mm. The pickup could become skewed relative to the track cable, but here it is assumed to be perfectly aligned (i.e., the vehicle is traveling in a straight line). The car should nominally drive in the center of the track; however, it is also desirable that sufficient power is available at the extremes to enable the car to return to the center of the track. A 3-D FEM analysis was undertaken using JMAG-Studio developed by The Japan Research Institute Limited. This work was undertaken prior to the actual construction of the track or

COVIC et al.: THREE-PHASE INDUCTIVE POWER TRANSFER SYSTEM FOR ROADWAY-POWERED VEHICLES 3375 Fig. 8. Measured track currents. (a) With current imbalances. (b) After balancing. Fig. 9. (a) Single ferrite. (b) Single winding double ferrite block. (c) Dual winding double ferrite block. TABLE I TRACK CONFIGURATIONS power supply. Three track designs were investigated, and their cable positions are given in Table I. Simulation results were undertaken using a similar approach to that discussed in [22] and later verified experimentally using the single pickup of Fig. 9(a). In both simulation and experiment, the pickup was shifted horizontally in 10-mm intervals from the track center. The track current was maintained at 40 A/phase in all cases, and the frequency of the supply was 38.4 khz. The open-circuit voltage and short-circuit current of the pickup as defined in (1) and (2) were obtained at each position. The uncompensated output power of the pickup was then calculated using (3). Results of this comparison are shown in Fig. 10. Fig. 10(a) presents the simulated results, whereas Fig. 10(b) shows the measured results. As noted, there is a close agreement, enabling further design and investigation using simulation. Small discrepancies result from slight variations in track Fig. 10. (a) Simulated S u versus distance off center. (b) Measured S u versus distance off center. positioning in the experimental system. As noted, Track 1 is not symmetrical. The phases closer to the track center are spaced closer together. As expected, the calculated uncompensated power is higher here and the power drops off rapidly. Tracks 2 and 3 are designed with equal spacing between phases, resulting in a more continuous power profile across the width of the track. A comparison of power profiles from these two configurations shows that the pickup width should be large enough to capture flux from more than one phase. Ideally, it should be at least twice the spacing between any two phases. In general terms, the power profile can be widened further if the spacing between the track phases is increased. However, comparing the results of Tracks 2 and 3, for a given pickup size, any increase in track spacing causes a corresponding (almost linear) reduction in the power density of the system, which

3376 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 54, NO. 6, DECEMBER 2007 Fig. 11. (a) Simulated S u versus distance for the single-coil pickup on a three-phase and a two-phase bipolar track. (b) Measured S u versus distance off center. reduces the uncompensated power available to the pickup (but without substantially affecting the system efficiency). The power peaks and troughs (notable as the pickup is moved across the complete width of the track) will also become more pronounced in a system with wider spaced tracks as the pickup is unable to capture all of the available field lines from neighboring track phases. A. Finalizing Track and Pickup Designs As shown, Track 3 has a power profile that extends nearly 135 mm but requires a wider pickup to ensure a greater proportion of the field flux is captured at all points across the track, thereby boosting power and providing a much flatter power curve over the center of the track. Consequently, Track 3 was chosen for the final design. JMAG was used to simulate the performance of a new improved double-width single-coil pickup, which was constructed as shown in Fig. 9(b) using two of the original ferrite blocks, on the chosen three-phase bipolar Track 3 layout. For the purpose of comparison, this new pickup was also simulated on a conventional single-phase bipolar track. Both tracks were assumed to have identical currents of 40 A/cable at 38.4 khz. The results are presented in Fig. 11(a). Maximum power is transferred when a vehicle s power pickup is perfectly aligned with the track cable. In practice, some misalignment from this ideal driving position is to be expected. In order to provide this tolerance in a single-phase track system, it is assumed that the target minimum power occurs at 50% of maximum so that a usable driving region is possible. Fig. 11(a) shows that a vehicle with a single pickup has two 30-mm-wide zones within which it can operate. Inside each driving zone the untuned pickup delivers a minimum of 3.4 VA. In comparison, the three-phase track delivers an untuned power of at least 7.5 VA over a 220-mm driving zone. As such, the three-phase track delivers a continuous power profile more than twice as large as the single-phase system, and that extends almost the entire width of the constructed track the lateral tolerance is at least three times better. In practice if the three-phase system is operated with slightly less than two thirds of the required track current/phase of the single-phase system, both systems will provide a similar average uncompensated power S u to a pickup operating in their respective usable operating regions. Thus, for a desired output power, the operational efficiency of both systems will be almost identical (approaching 80% 85% near rated operation) as the losses in the three-phase system running under this reduced track current are similar to the single-phase system despite the additional switches, transformers, capacitors, and track phases. The star delta transformer also tends to balance these losses across the inverter bridge irrespective of the position of the pickup. The designed prototype toy car system requires 30 W to be delivered from the power pickup. Using the three-phase track system with the vehicle driving near the track center, the uncompensated power achieved from Fig. 11(a) is approximately 9.5 VA so that the required Q 2 using (4) is around 3. This low circuit Q ensures a low voltampere rating for the pickup and high tolerance to tuning errors [15]. At 135-mm off track center, the measured uncompensated power is around 1.5 W. Assuming Q 2 =10is allowed by the vehicle power controller, this would enable sufficient power to drive the vehicle back toward the track center. Fig. 11(b) shows the resulting measured power profiles using two pickup designs constructed as shown in Fig. 9(b) and (c). Of note is that the power profiles of each are similar, with outputs close to that predicted using JMAG. The single-coil pickup is found in practice to have a smoother power profile over the center of the track and was consequently chosen in the final system. The profile of the single-coil system is surprisingly flatter than predicted by simulation [Fig. 11(a)]. This arises from slight practical variations in the constructed system compared with that simulated. The physical construction of the pickup of Fig. 9(b) required two ferrite blocks to be glued together. This process is not exact, and there exists a nonuniform air gap in construction. For the purpose of the simulation, this is assumed to be a uniform air gap of 0.4 mm. Furthermore, in practice, the coil width of the pickup is slightly larger (JMAG assumes a compact coil), and the track currents, while close, are not exactly 40 A/cable. A larger coil acts to reduce the secondary leakage and therefore improves coupling and power, while the presence of the air gap effectively lowers

COVIC et al.: THREE-PHASE INDUCTIVE POWER TRANSFER SYSTEM FOR ROADWAY-POWERED VEHICLES 3377 the relative permeability of the ferrite and therefore reduces power. Simulations have shown that if this air gap were not present, the power output would increase by more than 35%, which further enhances the benefit of the three-phase system over a traditional single-phase track system. VI. CONCLUSION A new three-phase IPT system has been developed. The power supply comprises a standard six-switch inverter with resonant LCL topology incorporating full isolation and protection means. A novel PWM generation scheme that enables full control of the track currents without modifying the dc voltage as would normally be required was presented. The effect of mutual inductance between parallel long cables results in imbalances between the track phases. However, by finely tuning the track inductance, the track can be balanced. The track geometry and pickup design were undertaken using FEM modeling and experimentally verified. A pickup having a width that couples flux from all three phases is desirable because of the time-varying nature of the field. The width of this pickup should be more than double the spacing between phases. The prototype system has shown that continuous power transfer to moving vehicles is possible without placing significant restrictions on vehicle operators. The prototype could easily be scaled to deliver the necessary power demands for practical electric vehicles. [12] F. Sato, J. Murakami, T. Suzuki, H. Matsuki, S. Kikuchi, K. Harakawa, H. Osada, and K. Seki, Contactless energy transmission to mobile loads by CPLS-test driving of an EV with starter batteries, IEEE Trans. Magn., vol. 33, pt. 2, no. 5, pp. 4203 4205, Sep. 1997. [13] F. Sato, H. Matsuki, S. Kikuchi, T. Seto, T. Satoh, H. Osada, and K. Seki, A new meander type contactless power transmission system Active excitation with a characteristics of coil shape, IEEE Trans. Magn., vol. 34, pt. 1, no. 4, pp. 2069 2071, Jul. 1998. [14] C.-S. Wang, O. H. Stielau, and G. A. Covic, Design considerations for a contactless electric vehicle battery charger, IEEE Trans. Ind. Electron., vol. 52, no. 5, pp. 1308 1314, Oct. 2005. [15] O. H. Stielau and G. A. Covic, Design of loosely coupled inductive power transfer systems, in Proc. Int. Conf. Power Syst. Technol. (Powercon), Perth, Australia, Dec. 4 7, 2000, vol. 2, pp. 85 90. [16] C.-S. Wang, G. A. Covic, and O. H. Stielau, Power transfer capability and bifurcation phenomena of loosely coupled inductive power transfer systems, IEEE Trans. Ind. Electron., vol. 51, no. 1, pp. 148 157, Feb. 2004. [17] C.-S. Wang, G. A. Covic, and O. H. Stielau, Investigating an LCL load resonant inverter for inductive power transfer applications, IEEE Trans. Power Electron., vol. 19, no. 4, pp. 995 1002, Jul. 2004. [18] M. Borage, S. Tiwari, and S. Kotaiah, Analysis and design of an LCL-T resonant converter as a constant-current power supply, IEEE Trans. Ind. Electron., vol. 52, no. 6, pp. 1547 1554, Dec. 2005. [19] B. Mangesh, T. Sunil, and K. Swarna, LCL-T resonant converter with clamp diodes: A novel constant-current power supply with inherent constant-voltage limit, IEEE Trans. Ind. Electron., vol. 54, no. 2, pp. 741 746, Apr. 2007. [20] B.-M. Song, R. Kratz, and S. Gurol, Contactless inductive power pickup system for Maglev applications, in Proc. Conf. 37th IAS Annu. Meeting, 2002, vol. 3, pp. 1586 1591. [21] S. Dieckerhoff, M. J. Ruan, and R. W. De Doncker, Design of an IGBT based LCL-resonant inverter for high-frequency induction heating, in Proc. Conf. IEEE IAS Annu. Meeting, Phoenix, AZ, Oct. 3 7, 1999, vol. 3, pp. 2039 2045. [22] D. Kacprzak, G. A. Covic, and J. T. Boys, An improved magnetic design for inductively coupled power transfer, in Proc. Conf. IPEC, Singapore, Nov. 1 Dec. 2005, pp. 1 4. REFERENCES [1] J. T. Boys, G. A. Covic, and A. W. Green, Stability and control of inductively coupled power transfer systems, Proc. Inst. Electr. Eng. Elect. Power Appl., vol. 147, no. 1, pp. 37 43, Jan. 2000. [2] J. Meins, F. Turki, and R. Czaniski, Contactless high power supply, in Proc. UEES, Crimea, Ukraine, Sep. 24 29, 2004, pp. 581 586. [3] J. M. Barnard, J. A. Ferreira, and J. D. van Wyk, Sliding transformers for linear contactless power delivery, IEEE Trans. Ind. Electron., vol.44, no. 6, pp. 774 779, Dec. 1997. [4] J. Lastowiecki and P. Staszewski, Sliding transformer with long magnetic circuit for contactless electrical energy delivery to mobile receivers, IEEE Trans. Ind. Electron., vol. 53, no. 6, pp. 1943 1948, Dec. 2006. [5] E. H. Lechner, D. M. Empey, and S. E. Schladover, Testing of a roadway powered electric vehicle prototype, in Proc. 10th Int. Elect. Vehicle Symp., Hong Kong, Dec. 3 5, 1990, pp. 959 973. [6] M. Eghtesadi, Inductive power transfer to an electric vehicle An analytical model, in Proc. 40th IEEE Veh. Technol. Conf., Orlando, FL, May 6 9, 1990, pp. 100 104. [7] K. W. Klontz, D. M. Divan, D. W. Novotny, and R. D. Lorenz, Contactless power delivery system for mining applications, IEEE Trans. Ind. Appl., vol. 31, no. 1, pp. 27 35, Jan./Feb. 1995. [8] T. Hata and T. Ohmae, Position detection method using induced voltage for battery charge on autonomous electric power supply system for vehicles, in Proc. 8th Int. Workshop AMC, Kawaski, Japan, Mar. 25 28, 2004, pp. 187 191. [9] H. Sakamoto, K. Harada, S. Washimiya, K. Takehara, Y. Matsuo, and F. Nakao, Large air-gap coupler for inductive charger, IEEE Trans. Magn., vol. 35, pt. 2, no. 5, pp. 3526 3528, Sep. 1999. [10] G. A. Covic, G. Elliott, O. H. Stielau, R. M. Green, and J. T. Boys, The design of a contact-less energy transfer system for a people mover system, in Proc. Int. Conf. Power Syst. Technol. (Powercon), Perth, Australia, Dec. 4 7, 2000, vol. 2, pp. 79 84. [11] G. A. J. Elliott, J. T. Boys, and A. W. Green, Magnetically coupled systems for power transfer to electric vehicles, in Proc. Int. Conf. Power Electron. and Drive Syst., Singapore, Feb. 21 24, 1995, vol. 2, pp. 797 801. Grant A. Covic (S 89 M 91 SM 04) received the B.E.(Hons.) and Ph.D. degrees from The University of Auckland, Auckland, New Zealand, in 1986 and 1993, respectively. He is a full-time Associate Professor with the Department of Electrical and Computer Engineering, The University of Auckland. His current research interests include power electronics, ac motor control, electric vehicle battery charging, and inductive (contactless) power transfer. John T. Boys received the Ph.D. degree from The University of Auckland, Auckland, New Zealand, in 1962. After gaining the Ph.D. degree, he worked for SPS Technologies, USA, for five years before returning to the academe. He is currently a Professor of electronics and the Head of the Department of Electrical and Computer Engineering, The University of Auckland. He is the holder of more than 20 patents. His fields of interests are motor control and inductive power transfer. Dr. Boys is a Fellow of the Institution of Professional Engineers New Zealand.

3378 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 54, NO. 6, DECEMBER 2007 MichaelL.G.Kissin(S 02) received the B.E. degree in electrical and electronic engineering from The University of Auckland, Auckland, New Zealand, in 2005. He is currently working toward the Ph.D. degree at The University of Auckland. His research interests include inductive power transfer and roadway-powered electric vehicles. Howard G. Lu received the B.Eng. degree from Guangdong University of Technology, Guangdong, China, in 1986 and the M.Sc.(Hons.) degree from the University of Waikato, Hamilton, New Zealand, in 1996. He then spent ten years as an Electrical Engineer at Guangdong Power Grid, China. He is currently a Technician of power electronics with the Department of Electrical and Computer Engineering, The University of Auckland, Auckland, New Zealand.

本文献由 学霸图书馆 - 文献云下载 收集自网络, 仅供学习交流使用 学霸图书馆 (www.xuebalib.com) 是一个 整合众多图书馆数据库资源, 提供一站式文献检索和下载服务 的 24 小时在线不限 IP 图书馆 图书馆致力于便利 促进学习与科研, 提供最强文献下载服务 图书馆导航 : 图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具