Carrier harmonic losses that consist of iron loss, eddy current loss, and ac copper loss are generated in a permanent magnet machine driven by a PWM inverter. It is already known that a higher motor inductance can lead to lower carrier harmonic losses. This paper investigated the carrier harmonic loss composition of two motors with identical dimensions but different inductances. The finite element analysis (FEA) results showed that the eddy current loss in the iron core accounts for most part of the carrier harmonic losses. It is also proved that the carrier harmonic loss of the iron core is quantifiable using theoretical calculations.
This paper presents a physics-based IGBT model and a junction temperature analysis of parallel-connected IGBTs under PWM operating conditions using a physics-based IGBT model. The authors developed a physics-based IGBT model in which the excess carrier distribution within a drift region is represented using a one-dimentional ambipolar diffusion equation. The physics-based IGBT model makes it possible to predict losses and switching waveforms for converter applications. IGBTs are connected in parallel for medium- or large-capacity converters. In these applications, a transient current imbalance might occur owing to a difference in the wiring inductance or device characteristics between the IGBTs. An experiment shows that a difference in the wiring inductance between the two IGBTs causes a transient current imbalance, and the result is in excellent agreement with the result of a simulation using the physics-based IGBT model. The two parallel-connected IGBTs in this study correspond to a power module for 3.7kW motor drives, and the junction temperatures of both IGBTs are simulated by electro-thermal simulation under the following two conditions: a difference in the wiring inductance and a difference in the device characteristics. The temperature difference between the two IGBTs is approximately 4-7℃ under the applied conditions: a wiring inductance mismatch (20nH) and a threshold voltage mismatch (0.5V). The validity of the physics-based IGBT model is verified, and the model is found to be very useful when designing power converters.
The authors have proposed a new method to decrease high-frequency harmonics at a specific frequency band by modifying the switching transient slope. In previous studies, there were several problems in applying modified transient pulse width modulation (MT-PWM) to actual converters. In this paper, three problems are solved using an improved MT-PWM method. First, the MT-PWM signal was obtained with a trial-and-error method that involved complex computation procedures in the previous studies. In this paper, a new calculation procedure for obtaining the MT-PWM waveform with a simple calculation is proposed. Second, a new VDS controller (drain-source voltage controller) based on voltage feedback is proposed to realize a modified switching transient to increase the stability of the switching operation. Third, the dependency of MT-PWM on source voltage variation is investigated in order to implement MT-PWM in an actual step-down converter. From this result, the concept of a new type of VDS controller considering the source voltage variation is proposed. Finally, the authors attempted to apply MT-PWM to an H-bridge converter to expand the application of MT-PWM. An H-bridge converter with MT-PWM for a DC motor drive is successfully operated in the experiment.
A critical issue for building air conditioning is shifting and cutting peaks in electricity demand. Appropriate measures should consider not only the heat resource supplier side but also the user side. Demand and peak cut are controlled from the supplier side. This study is focused on the user side because conformity and energy conservation should also be pursued at the same time. This paper proposes a user-driven heat resource allocation method for client-server building air conditioning systems. Each client represents a room or residential area that requires heat resources. The server is the heat resource supplier. The heat resource allocation problem model was formulated as server-client architecture and solved as a 0-1 integer programming problem. The proposed method enabled the peak heat resource use of each room to be shifted while maintaining the residential environment. The effectiveness of this method was proven through simulation.
The authors previously developed a simplified whole-body haptic system named “Haptic Armor” that consists of a force sensor device and a convex-shaped end-effector. However, this system suffered from the limitation that the shape of the end-effector had to be convex. Therefore, this paper proposes a whole-body haptics method based on a force sensor device and a non-convex-shaped end-effector. First, this method estimates candidate contact points from the force sensor information. Then, it identifies the contact point by using the information on the variation in the external force direction while the end-effector is in contact.
This paper demonstrates a control method that can reduce the start-up time of a weaving machine based on vector control. Traditionally, the induction motor used for a weaving machine was driven by a direct power grid connection, that can achieve a quick start-up time for the weaving machine. However, it is inefficient in steady-state operation. An inverter is utilized in the induction motor to obtain a higher efficiency than that of the induction motor directly connected to the power grid. However, the driving of the inverter slows down the start-up time of the weaving machine. This paper proposes a fast accelerating method for the weaving machine drive system, in which an open winding of an induction motor is connected in series to an inverter and delta-star switching unit. From the results, the start-up time of the proposed system is shown to be the same as that of the direct power grid connection. In addition, the inrush current of the proposed system is decreased by 37.7% compared to that of the direct power grid connection.
This paper presents a high-precision sensorless force control for positioning devices with contact operation. The sensorless force control is designed with a sliding mode controller and contact model that provide the required control specifications to compensate for the non-linear spring characteristics of the contact mechanism. The effectiveness of the proposed control approach was verified through numerical simulations and experiments using a prototype.
This paper describes a 750-V, 100-kW, 20-kHz bidirectional isolated dc/dc converter with the “dual-active-bridge” configuration using four SiC-MOSFET/SBD two-in-one modules. It pays particular attention to conversion efficiency and power-loss breakdown. The overall maximum efficiency from the dc-input to the dc-output terminals is measured accurately to be 98.7%, excluding the gate-drive and control circuit losses from the overall power loss. A power-loss breakdown at the rated (100kW) operation is carried out to separate conduction, switching, iron, copper, and unknown losses from the overall loss. The power loss breakdown indicates that the sum of the conduction and switching losses produced in the four SiC modules is about 60% of the overall loss, and that the conduction loss is almost equal to the switching loss. Moreover, this paper provides theoretical and experimental discussions on a range of dead times for achieving zero-voltage switching (ZVS).
This paper proposes a linear PFC regulator for LED lighting applications. The proposed circuit is small in size because the circuit structure consists of only semiconductors, ICs, and small resistors without any reactors and electrolytic capacitors. The current bypass circuit which is connected in parallel to the LED string consists of one MOSFET, two Zener diodes, and one resistor. The MOSFET is operated in an active state by a self-bias circuit. Thus, an external controller and high voltage gate drivers are not required. The proposed circuit is experimentally validated using 7.8W 5-string and 15W 10-string prototypes. From the experimental results with the 10-string prototype, the total harmonic distortion of the input current is 5.1%, and the power factor is 0.999. In addition, the power loss is analyzed at an efficiency of 91.6% for the prototype circuit.
The phenomenon of cross coupling between coils is a widely discussed subject in various fields. In communication technology, for example, the resonator is used as a filter. Cross coupling affects the filter characteristic and therefore should be eliminated. The earlier studies, however, are either not applicable to wireless power transfer or lack a detailed treatment of wireless power transfer efficiency. Papers that do mention the undesirable effects of cross coupling on wireless power transfer characteristics do not provide an analysis of how cross coupling affects the transfer efficiency. In this study, we analyzed the effects of cross coupling on the overall efficiency, with a detailed comparison of the results with and without cross coupling. We discovered that at the resonant frequency, as the cross coupling increases, the frequency shift also increases, which causes the efficiency to deteriorate. From this result, we proposed four ideas for efficiency improvement. Among them, the frequency tracking method is most useful for efficiency improvement when the operating frequency is variable. When the operating frequency is fixed, the cross-coupling canceling method can be used to achieve maximum efficiency at the resonant frequency. We also discovered that load value optimization is not needed in the frequency tracking method because the simple tracking method can provide sufficient efficiency improvement.
This letter proposes a new wind power generator system using a three-phase voltage-doubling rectifier that consists of only passive elements. The proposed generator system can obtain maximum power output from wind turbines without using the control circuit. The effectiveness of the proposed generator system is demonstrated with simulations and experiments.
This paper presents a calculation reduction method for Laguerre functions used in identification. The conventional calculation method uses repeated calculations for the summation and optimization. The matrix representation of the Laguerre function and its coefficients is proposed to solve this problem. In the proposed method, repeated calculations are not required because the Laguerre functions and coefficients are described using the product of the coefficient and variable matrices. Simulation results show that the proposed method is effective for calculating Laguerre functions and coefficients.