Your Paper's Title Starts Here - Universiti Putra Malaysiasaes2017.upm.edu.my/ENG Poster...

5th International Symposium on Applied Engineering and Sciences (SAES2017)

14th–15th November 2017 | MALAYSIA

UNIVERSITI PUTRA MALAYSIA, SERDANG, SELANGOR

1

Poster code: E1

Enzymatic Hydrolysis of N.gaditana sp for Bioactive Peptide Production

Nur Izzati Md Saleh1,*, Wan Azlina Wan Ab Karim Ghani1, Razif Harun1, And Siti Mazlina Mustapa Kamal2

1Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, Malaysia

2 Department of Food and Processing Engineering, Faculty of Engineering, Universiti Putra Malaysia, Malaysia

*Corresponding author’s e-mail: [email protected],

Abstract. Microalgae Nannochloropsis gaditana (N.gaditana sp) was used as a starting material to produce microalgae protein hydrolysate (MPH) by protease enzyme. The effect of pH, temperature, enzyme concentration and substrate concentration on the degree hydrolysis (DH) were investigated. The screening of enzymatic hydrolysis reaction time was conducted from 1 hr and up to 48 hours. The enzymatic hydrolysis reactions were conducted for 24 hours for all set experiments based on the screening time. The maximum cleavage of peptide bonds occurred was found at pH 8 (DH=41.69%), temperature 55°C (DH=42.53%), enzyme concentration 0.3 mg/L (DH=43.96%) and substrate concentration 2 g/L (DH=58.70%). These results demonstrated that N.gaditana sp is rich in bioactive peptide which has the capability to be an antioxidant.

Keywords: Enzymatic hydrolysis, bioactive peptide, microalgae, degree hydrolysis, antioxidant




2

Poster code: E3

Coal Bottom Ash for Palm Oil Mill Effluent (POME) Decolourization

Syahin Saleh1, Wan Azlina Wan Abd Karim Ghani2,* and Loh Soh Kheang2 1 Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra

Malaysia, 43400 UPM Serdang, Selangor, Malaysia 2 Energy and Environment Unit Engineering and Processing Research Division Malaysian Palm Oil

Board, Bandar Baru Bangi, 43000 Kajang, Selangor, Malaysia

*Corresponding author’s e-mail: [email protected]

Abstract. The utilization of coal bottom ash (CBA) from thermal power plant for palm oil mill effluent (POME) decolourization was studied. CBA is prepared using chemical activation with hydrochloric acid (HCl) to increase adsorption performance. Physico-chemical characterizations of CBA-HCl was analysed using BET for surface area, FTIR for its surface chemistry, SEM for morphology and EDX for elemental analysis. The effectiveness of CBA-HCl for colour adsorption of POME was investigated as a function of pH, initial concentration, adsorbent dosage and contact time by batch experiments. Adsorption increased with increasing contact time and the equilibrium states could be achieved in range of 18 to 24 hr. Results showed that CBA-HCl perform with maximum colour removal of 93% at pH 6 with 10% (w/v) adsorbent dosage in 24 hr. It was shown that the isotherm for adsorption of colour from POME on CBA-HCl was well fitted by Freundlich equation (correlation coefficient, R2 = 0.9636). Based on the results, CBA showed potential adsorbent candidate for POME decolourization.

Keywords: POME, Decolourization, Colour, Removal, Adsorption, Bottom ash.




3

Poster code: E4

The influence of Iteration Number in PSO application for Research

reactor TRIGA PUSPATI

Nor Arymaswati Abdullah1, 2*, Azura Che Soh1, Ribhan Zafira Abdul Rahman1, Samsul Bahari Mohd Noor1and Julia Abd Karim2

1Faculty of Engineering, Universiti Putra Malaysia, 43400 Serdang, Selangor Malaysia 2Technical Support Division, Malaysian Nuclear Agency, Bangi 43000 Kajang, Selangor Malaysia


Abstract. Global best position is one of the components in Particle Swarm Optimization to determine the ultimate result among the best solutions with the smallest error. This paper describes how the iteration affected the global best value and performs different output of the research reactor TRIGA PUSPATI. The swarm number is the same for both cases, but the number of iterations are 20 and 10 were studied. Generally, different iteration number gives a different value of global best value which yields different proportional gain value. The studies found that, when these different proportional gain values were applied into the system, the power output showed that the differences between them are small.

Keywords: Global best position, iteration number, proportional gain




4

Poster code: E5

Rank-based Optimal Neural Network Architecture for Dissolved Oxygen

Prediction in a 200L Bioreactor

Nor Hana Mamat1,2 *, Samsul Bahari Mohd Noor1,, Azura Che Soh1, Farah

Saleena Taip1,, Ahmad Hazri Ab. Rashid3,, Nur Liyana Jufika Ahmad3,, and Ishak Mohd Yusuff3,

1Faculty of Engineering, University Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia 2Faculty of Electrical and Automation Engineering Technology, TATI University College,

Teluk Kalong, 24000 Kemaman, Terengganu, Malaysia 3Industrial Biotechnology Research Centre, SIRIM Berhad, Persiaran Dato Menteri,

Seksyen 2, 40700 Shah Alam, Selangor, Malaysia

*Corresponding author’s e-mail: norhana.mamat @tatiuc.edu.my

Abstract. In a fermentation process, dissolved oxygen (DO) concentration is mostly affected by aeration rate, and agitation speed and temperature. Thus it is beneficial to model the relationship of DO concentration with these variables based on real process data for further use in controller design. Formulation of bioprocess model using process data or data driven technique is able to describe the true process conditions better than a model driven technique that focused on ideal steady state condition of process map the relationship of DO concentration with other physical and chemical process variable that has influence on the process. Artificial neural network (ANN) is a reliable and popular tool for approximation of nonlinear relationship between input and output data with little knowledge and no assumption of the process, also when dealing with problems involving prediction of variables. The structure of a neural network model namely input layer, hidden layer and output layers has significant effect on predicted results. While the number of neurons in input and output layers are determined based on the number of respective input and output parameters, there is no straightforward method to determine the optimal number of neurons in hidden layer. In order to select the appropriate structure, trial and error method or repeated runs are usually used to find the number of hidden neurons that gives smallest value of error and highest value of correlation coefficient. In this paper, a ranking system based on repeated runs of neural network model is used to determine the architecture with optimal number of hidden neurons for three different division of data for training and testing. The ranks are applied together for both training and testing datasets. The backpropagation neural network model with Lavenberg Marquardt learning algorithm was developed using 1476 samples real process dataset obtained from a fermentation process in a 200L bioreactor. The ranking system applied to simulation results shows that the best prediction of dissolved oxygen level was obtained for 80%/20% data division with 6 hidden neurons.

Keywords: Neural network, optimal architecture, bioreactor, dissolved oxygen




5

Poster code: E6

Study on the improvement of the Total Harmonic Distortion Using

Gradation Multi-Level Inverter

Takaya Masuzaki1, Tsuyoshi Hanamoto Department of Biological Functions and Engineering, Graduate School of Life Science and

Systems Engineering, Kyusyu Institute of Technology, 2-4, Hibikino, Wakamatsu, Kitakyushu, 808-0196, Japan


Abstract. Recently, various applications using motors with inverter have been progressed where the inverter is a DC/AC conversion equipment for efficiently controlling a motor, and achieves speed control of the AC motor to convert the amplitude and frequency of the applied voltage. The inverter can convert the power efficiently using power semiconductor as an on-off switch. But the output voltage of the inverter takes just three level (E, 0, -E) for ordinary voltage souse inverter (VSI). Therefore, the multi-level inverter is introduced increasing output voltage level. A multi-level inverter can increase the resolution of the output voltage by increasing series connected full bridge single phase inverter. For examples, some kind of power conditioners and uninterruptible power supply employing the multi-level inverter. In this research, instead of the conventional multi-level inverters control method, different value of the voltage source is used for the power supply called gradation type is adopted. Even if the same number of the voltage source is connected, the gradation type multi-level inverter has an advantage that a larger number of voltage patterns can be obtained compared with the conventional multi-level inverter. For example, the gradation type multi-level inverter with 1:3:9 the ratio of power supply can obtain 27 voltage level. Both simulation results and experimental results shows the validity of the gradation type multi-level inverter to measure and compare the total harmonic distortion (THD). From the result, the usefulness of the new method is confirmed by grasping whether THD can be reduced.




6

Poster code: E7

Improvement of the Development Environment Using Hardware CO -

Simulation and IP Catalog

H. Ahmad1, M. Nizam1, N. Effendy1, I. Fakhrul1, A. Tanaka2 1Department of Electrical & Electronic, Faculty of Engineering, University Putra Malaysia, 43400

UPM Serdang, Selangor, Malaysia 2Faculty of Engineering, Shinshu University, 4-17-1 Wakasato, Nagano, 380-8553, Japan


Abstract. In this paper, improvement of the development environment using hardware CO - Simulation and IP Catalog of MATLAB / Simulink for the control of matrix converters is shown. In recent years, environmental problems are attracting attention, and the study of renewable energy and the stability analysis of these energies to the grid connection is discussed vigorously. Matrix converters have been researched as one candidate for the next generation power converters. Because the matrix converter does not have a DC link and directly converts AC power into AC power of arbitrary frequency. The purpose of this study is a digital hardware control of 3-phase to 3-phase matrix converter according to the calculations for generating DDPWM (direct duty ratio PWM) using the FPGA. After designing the block diagram with MATLAB / Simulink and confirming the simulation result, we convert the block by IP Catalog using System Generator and implement it on the real machine. However, there is an issue that the simulation time takes so long because the simulation is done every step in the software. Therefore, by using Hardware CO - Simulation, it is possible to reduce the time to use the control block made by System Generator sent the FPGA device. The validity of this study is confirmed that the output of the hardware CO-Simulation agrees with the experimental result. In addition, by using IP Catalog, convenience is improved because the GUI based project can be possible instead of the hardware description language, called VHDL, based project.




7

Poster code: E8

Constructing of Efficient Development Environment of Motor Control

System Using GUI System

M.Ono, T. Hanamoto Kyushu Institute of Technology, 2-4 Hibikino, Wakamatsu, Kitakyusyu, Fukuoka, 808-0196, Japan


Abstract. At the present, mminiaturization and high efficiency of motor are being advanced due to progress of technology. Along with these, the control of the electric motor becomes complicated, and the control program capacity has become larger. The conventional CPU control system is sequential processing, and when large capacity processing is performed, many delays occur and securing real time property becomes difficult. One of the solution is FPGA. FPGA is a hardware element capable of rewriting a circuit. However, FPGA don’t have the flexibility like CPU. Furthermore, In the FPGA, it takes a lot of time to change the circuit, for example, it is necessary to rewrite the circuit again when we change some parameters and some part of circuits. Hence, in my research, we develop a motor control system development environment using System Generator which is the block set of Simulink. Using System Generator, it is possible to create a circuit of FPGA using GUI system so the knowledge of the conventional hardware description language writing code is no required. Then it is possible to change the circuit more visually and efficiently as compared with the conventional work of rewriting the code when changing the circuit. And in the system generator, control blocks for example, speed PI control, current PI control, etc. can be created as individual control blocks. As a result, it is considered various control systems that is using an electronic motor can be constructed more efficiently by diverting these blocks.




8

Poster code: E9

Establishment of a Development Environment for Motor Control

Systems Using a Multifunctional FPGA

R.Hashimoto, M.Ono, T.Hanamoto Department of Biological Functions and Engineering, Graduate School of Life Science and

Systems Engineering, Kyushu Institute of Technology, 2-4, Hibikino, Wakamatsu, Kitakyushu, 808-0196, Japan


Abstract. This paper presents a development environment for permanent magnet synchronous motors (PMSMs) control system using a MPU core on the multifunctional FPGA. In recent years, demands for technologies such as variable speed driving and high accuracy of motor control are increasing according to progress of power electronics technology. As a result, control computational algorithm becomes complicated and the capacity of the program is increased. Therefore, it is difficult to ensure real-time performance by controlling with the conventional MPU. In this study, we construct a real-time control system of PMSMs including complicated control operation using a MPU core on the multifunctional FPGA. Usually, PMSMs system controlled by software is difficult for high frequency real-time control because of a sequential processing. Then, real-time control of PMSMs by using FPGA, which is possible to operate with high frequency by parallel processing, is proposed in this paper. Currently logic based FPGA system for the speed control of PMSMs exists, however, there is a limitation to treat the various input and output parameters flexibly. So in this paper, it carried out construction of the development environment of PMSMs speed control system using a MPU core on the multifunctional FPGA. The current progress is as follows, it is realized changing the speed command value and the gain value from the MPU to FPGA and reading the speed calculation value from FPGA to MPU online without connection to PC.




9

Poster code: E10

Two-Degrees-of-Freedom Disturbance Suppression

Speed Controller Design Using E-FRIT

J. Fukushima, Y. Takahashi , T. Hanamoto Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology, 2-4

Hibikino, Wakamatsu, Kitakyushu, Fukuoka, 808-0196, Japan


Abstract. This paper presents a new speed controller design method of PMSM (Permanent Magnet Synchronous Motor) using E-FRIT (Extended Fictitious Reference Iterative Tuning) to design both the response of the speed command and disturbance response. To get the desired response of the system, the modelling of the system is required and it needs system parameters and the system structure. Also, it is necessary to repeat experiments to get the desired response. It is waste of the cost and time. Accordingly, controller design using E-FRIT is proposed in this paper. E-FRIT is the method to tune the parameters of a controller offline based on the input/output data get only one time closed loop control experiment. E-FRIT is possible to perform the design of the optimal target transfer function for the system, and the same time to adjust the controller parameters. Here I consider a PI-P speed controller as a simple Two-Degrees-of-Freedom controller, and parameters of the controller are designed for both step response and disturbance response at the same time of PMSM using E-FRIT. The proposed method features the controller parameters are not designed separately for speed and disturbance responses. The proposed method is considered the signal of disturbance moves virtually to the speed reference signal. The design of the disturbance response is possible to evaluate the speed response equivalently. Consequently, PMSM speed control system, where the desired response for both the step response and disturbance response, is achieved. The validity of the proposed method is shown in the experiments.




10

Poster code: E11

Study on Nonstationary Noise Removal for AE Sensing

Yuya Akimoto Department of Human Intelligence Systems, Graduate School of Life Science and Systems

Engineering, Kyushu Institute of Technology, 2-4, Hibikino, Wakamatsu, kitakyushu, Hukuoka, 808-0196, Japan.


Abstract. This paper presents, we describe a method of removing nonstationary noise generated by using AE sensor which is one of non-destructive inspections. Test method using AE (Acoustic Emission) method as one of non-destructive inspections have been used. AE is stored until deformation or destruction inside the object the strain energy is released and a part thereof is radiated as an elastic wave (AE wave) in addition, AE ranges from several kHz to several MHz, the frequency band is wide, the AE wave is small even in case of initial failure such as occurrence of friction due to poor lubrication it is easy to find. The AE method mainly detects the signal in the ultrasonic region and detects the AE wave emitted by the object itself. In this research, we are targeting machine processing machines as one application example of AE sensor. By using AE sensor for a machine tool, it was difficult to obtain in the past we believe that it is possible to acquire minute signals of abnormality. However, in the actual machining section, it is thought that mixing of noise above is considered, and thinking that a more precise noise removal method is necessary it can be selected. For the development of elemental technology of noise removal method in AE sensing, we aim to construct a noise canceler by adaptive filter based on independent component analysis and compare it with conventional adaptive filter and evaluate it.




11

Poster code: E12

Resolved Acceleration Control of a Underwater Robot Holding a Fixed

Object Underwater with One Hand

K.Hieshima Department of Human Intelligence Systems, Kyushu Institute of Technology,Kitakyushu, Fukuoka,

808-0135


Abstract. For various purposes, ocean development is being carried out. But, the work in the sea is dangerous for a human being. In regard, we develop an underwater robot which works in place of a person. This robot has two manipulators which have three links each. We control an underwater robot in “Resolved Acceleration Control”. This method is the feedback control law that can control a vehicle and manipulators at the same time, but this method is applied to only the robot which is in complete floating state. When the robot is not the state, I simulated and checked whether the control method could apply to it. I assumed a state catching a fixed object in the water by one arm (armL), and controled other arm (armR). To fix a position of armL’s hand, I

made the following program. If the position of the hand changes, the program produces power F which the program returns to the original position on the hand. As a result of simulation, although halfway the error was delivered to the tolerance level, the position of a vehicle and the hands did not converge from a certain point in time. In regard, the result ended in failure. As consideration, from the mechanic condition of the robot, I think I did not set the value of a feedback gain applied to a vehicle and an arm definitely.




12

Poster code: E13

Study of the Control Method of the Microrobot for Competitions

S.Kato1

1Kyushu Institute of Technology Graduate School of Life Science and Systems Engineering


Abstract. In late years, with various innovation, the research and development of the robot advance. Above all, the study on downsizing of the robot is conducted flourishingly. A study has been pushed forward as the purpose that worked at a small place and the dangerous place that the microrobot made use of an advantage of the size, and a person did not put. Of the microrobot which was equal to the rule of the sumo micro mechanism section of the international micromechanism contest in this study produce it, and control it. I designed it on a PC and produced it by manual labor. The materials of the main body of robot assumed it aluminum which the processing was easy to make, and the parts such as wheels used brass, brass. A competition board was small and I often fell from a competition board before jostling with a partner by an operation error by the vehicle driver because the motor of the robot worked quickly and was seen. Therefore I decided to produce the robot which I prevented an operation error by the human being by controlling it, and did not fall from the competition board using the sensor. Because there are many improvement of the control method when an anteroposterior sensor reacted as a future problem, control of the number of revolutions of the motor, wiring, it is necessary for the laborer such as reducing number to think.




13

Poster code: E14

Study on Position Sensorless Starting Method for Surface Permanent

Magnet Synchronous Motor (SPMSM)

K.Maeda, T. Hanamoto Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology, 2-4

Hibikino, Wakamatsu, Kitakyushu, Fukuoka, 808-0196, Japan


Abstract. In this paper, sensorless start-up method for surface permanent magnetic synchronous motors (SPMSMs) is proposed. Driving and automation using motors are growing rapidly and becoming more-complicated due to large-scale industrial application. The emphasis on high efficiency control of motor with lower cost is needed for modern industrial processes and product development. Among the motors, SPMSMs have been widely used in industrial applications due to its high power and efficiency capabilities. To achieve high performance control of SPMSM, the rotor position is required to be detected precisely. In general, position detector such as an encoder or a hall sensor has been employed. The position detectors are having some demerits such as high cost, degradation, upsizing, and less reliability. To overcome these demerits sensorless control of SPMSM can be adopted. In the previous work for sensor less starting of SPMSMs, DC exciting current flows to the motor for vibrating and rotor position was estimated from the measured induced voltage at the start up. In that case, the starting control was successfully achieved by combining two observers, a nonlinear observer and an induced voltage observer. However, that method has a disadvantage that rotor rotate the reverse direction for a short time in certain initial rotor position at the start-up. In the proposed method, to measure the time when the error between the induced voltage and the estimated induced voltage gets within a certain value, it is possible to rotate the right direction without reversing at the ± 90 ° initial position.




14

Poster code: E15

Device Destruction during AC Stress of a Thin-Film SOI Power N-

MOSFET

M. Nomura, A. Watanabe, S. Matsumoto Kyushu Institute of Technology, Sensui-cho, Tobata-ku, Kitakyushu-shi, Fukuoka, 804-8550,

Japan


Abstract. Silicon on insulator (SOI) technology has attracted attentions for high temperature applications because it can reduce leakage current and suppress the latch-up at high temperature [1]. In power ICs, power MOSFETs usually used of dynamic operation, in this paper, we explore device destruction phenomenon during the AC stress of thin-film SOI nMOSFET. Table 1 shows the mapping of the device destruction or not during stress. Stress time is 3600 seconds, and stress drain voltage is 6V. We use square wave (0V-4.5V) at AC stress. The device destruction occurs when the switching frequency of 1 kHz and 10 kHz. The power MOSETs do not break when the frequency is more than 100 kHz at temperature of less than 473 K. These are caused by accumulation of a minority carrier and this enhances the parasitic bipolar effect [2]. The parasitic bipolar effect causes by the low gate voltage of the rise and fall time of the pulse. Therefore, DC stress does not occur device destruction. The destruction occurs even at 100 kHz when temperature is 573 K. Then, we simulated the body hole current using TCAD. Dependence of body hole current on time is shown in Fig. 1. As temperature increases, body hole current tends to increase because leakage current at off state become larger at 573 K and this enhances the parasitic bipolar effect. We investigated device destruction phenomenon during the AC stress of thin-film SOI nMOSFET in high temperature. Destruction in low frequencies is caused by accumulation of minority carriers and enhance the parasitic bipolar effect.




15

Poster code: E16

DC Hot-carrier Effects in P-Channel Thin-Film SOI Power MOSFET

K. Ikeno, S. Matsumoto Kyushu Institute of Technology, Sensui-cho, Tobata-ku, Kitakyushu-shi, Fukuoka, 804-8550,

Japan


Abstract. Power ICs using Silicon-on-insulator (SOI) technology for high frequency switching and wireless communication applications has become attractive because of lower parasitic capacitance, completely isolating the devices each other and operating at high temperature 1). The thin-film approach has been quite promising because it can minimize the parasitic capacitance and reduce thermally induced leakage current 2). In power ICs, both n- and p- channel power MOSFET usually used and the reliability is important. We describe the hot carrier effect of the p-channel SOI power MOSFET. Dependence of threshold voltage shift on the stress gate voltage is shown in Fig. 1. The comparison of the threshold voltage shift for pMOSFET with that of the nMOSFET is shown in Fig. 2. For pMOSFET, threshold voltage shift goes plus direction when the gate voltage is more than -4 V and it goes minus direction when the gate voltage is less than -4 V. These results mean that hot electrons are injected into gate oxide when the gate voltage is more than -4 V and hot holes are injected into gate oxide when the gate voltage is less than -4 V. For nMOSFET, threshold voltage shift increase with increasing gate voltage and only hot electrons are injected when the gate voltage is more than (Vth+0.2) V 3). We clarify that minority carriers are injected when the gate voltage is low for pMOSFET. That is electrons or holes are injected depending on a stress gate bias condition for p-channel MOSFET.

mailto:[email protected]




16

Poster code: E17

Enhancement of the Thermal Properties of Suspended Glass Thin Films

Due to Long Range Surface Phonon-Polaritons

S. Hamamura1, L. Tranchant1, J. Ordonez-Miranda2, S. Volz3,4, K. Miyazaki1 1Department of Mechanical and Control Engineering, Kyushu Institute of Technology, 1-1 Sensui-

cho, Tobata-ku, Kitakyushu 804-8550, Japan 2Institut Pprime, CNRS, Université de Poitiers-ENSMA, 2 Rue Pierre Brousse, Bâtiment B25, TSA

41105, 86073 Poitiers Cedex 9, France 3Laboratoire d’Énergétique Moléculaire et Macroscopique, Combustion, UPR CNRS 288, École

Centrale Paris, Grande Voie des Vignes, 92295 Châtenay-Malabry, France 4LIMMS - CNRS-IIS (UMI2820), University of Tokyo, 4-6-1 Komaba, Meguro-ku, Tokyo 153-8505

Japan


Abstract. Surface phonon-polaritons (SPhPs) are evanescent electromagnetic waves propagating at the surface of dielectric polar materials such as glass. Their wavelengths belong to the mid-infrared and especially around 10 m for glass which make them potential powerful heat carriers at the nanoscale. The thermal performances of nanodevices decrease indeed with the size, this phenomemon has strong impact in the microelectronics industry having difficulties of heat management for electronic nanocomponents. Recent studies concerning glass thin films make no exception with that: the measured cross-plane thermal conductivity of silica films on a silicon substrate shows a decrease with the thickness. However the power of SPhPs have not yet been exploited properly experimentally. Theoretical studies have proved that the in-plane thermal conductivity of amorphous silica thin films can increase dramatically for thicknesses lower than 200 nm if they are suspended. This work aims at proving experimentally this fact. Suspended amorphous silica membranes have been fabricated by dry oxidation with various thicknesses from 75 to 300 nm. Then we measured the in-plane thermal conductivity by using the AC calorimetry method. Our preliminary results show thermal conductivities of 1.43 W/(m.K), 1.62 W/(m.K) and 1.71 W/(m.K) at room temperature for the samples of thicknesses 288 nm, 135 nm and 77.5 nm respectively. These results are very encouraging and consistent with the bulk glass thermal conductivity and the theoretical predictions. In future experiments we plan to confirm this tendency by measuring the thermal conductivity for different thicknesses and to rise the samples temperature for enhancing the surface effects.




17

Poster code: E18

An Investigation of a Three-Dimensional Stacked Power Supply on Chip

for High Frequency Switching Applications

Kengo Hiura, Kouta Ono, and Satoshi Matsumoto Kyushu Institute of Technology, Kitakyushu 804-8550, Japan


Abstract. The power supply on chip (SoC), which integrates passive components such as inductors and capacitors with semiconductor devices, has been attracting attention because it can realize the ultimate miniaturization of the power supply. A three-dimensional (3D) stacked power SoC (Fig. 1), which integrates GaN power devices, Si-LSIs, and passive devices was proposed [1]. It reduces parasitic capacitance and inductance markedly, thus it can operate high efficiency at high frequency. In this paper, we estimate the 3D stacked power SoC using circuit and electro magnetic simulations. We simulated buck converter (Vin =5.0V, Vout=2.5V) [2]. The maximum efficiency is 88.6% at 0.32A and 100MHz. In this case, power density is 404W/cm3. The efficiency increases with decreasing switching frequency at a higher output current because the influence of the resistance of the inductor is large at a higher current and the resistance increases as the frequency increases. We analyzed loss breakdown at 100MHz and 200MHz. The switching loss at 200MHz is higher than that at 100MHz. According to the simulation, buck converter converts voltage at high efficiency from 100MHz to 150MHz. We predicted the 3D stacked power SoC for a high-frequency DC-DC converter by numerical simulations. The efficiency of the 3D stacked power SoC is estimated to be 88.6% at 0.32A and 100MHz, and a chip size of 2 x 2 mm2.

Fig. 1 3D stacked power SoC Fig. 2 Dependence of the output current on the efficiency

Fig. 3 Loss Breakdown

References: [1] Kengo Hiura, et al., JJAP, 2017, 56, 04CR13 [2] D. Reusch, Ph.D. Thesis, Virginia

Polytechnic Institute and State University (2012).




18

Poster code: E19

Development of Personal Mobility That Enables Operation by Physical

Action

A.Kamikawatoko, H.Miyamoto Department of Human Intelligence Systems, Graduate School of Life Science and Systems

Engineering, Kyushu Institute of Technology, 2-4, Hibikino, Wakamatsu, Kitakyushu, 808-0196, Japan


Abstract. In this paper, we propose a prototype of personal mobility that enables operation by physical action for indoor light work. Today, considering the situation of Japanese houses, it is difficult for conventional wheelchairs to move a physically disabled person in a narrow indoor space. As a solution to support movements in such environments, we recommend that the wheelchair move freely and that the wheelchair be downsized. Therefore, this laboratory is developing the omnidirectional wheelchair using spherical wheels, and it is now possible to move to all directions. For this reason, the problem of miniaturization of the wheelchair has been changed from the wheelchair to the chair sitting in a standing position and the operation by the joystick has been improved to the operation by physical action. Rotational motion was made possible by using the difference between the measured distances of the two distance sensors, and parallel motion was made possible by using acceleration sensor. We have developed a new prototype of small personal mobility of omnidirectional movement mechanism by using spherical wheel. Compared with conventional operation, it is possible to operate by physical action rather than operation by joystick, so there is the advantage that you can work freely with both hands. It shows the performance equivalent to that of conventional prototype in the experiments.




19

Poster code: E20

Disturbance Response Design of 3 Inertia Resonance System Using

FRIT

Y. Takahashi, J.Fukushima, T.Hanamoto, M.Mohri Graduate school of Life Science and Systems Engineering, Kyushu Institute of Technology, 2-4

Hibikino Wakamatsu, Kitakyushu, Fukuoka, 808-0196, Japan


Abstract. This paper presents a disturbance response design of 3 inertia resonance system using FRIT (Fictitious Reference Iterative Tuning). When the motor and load are connected by a flexible shaft, it becomes a resonance system and the vibration occurs for adding the load or changing the command value. Then, it is necessary to treat it as a multi inertial system to which multiple loads are connected. To get the desired response of the system, the modelling of the system is required and it needs system parameters and the system structure in general. Also, it is necessary to repeat experiments to get the desired response. FRIT means fictitious reference iterative tuning and is a method of offline based performing parameter adjustment of the controller with only one time input and output data obtained by the closed loop response. FRIT can design the optimum response of the system. The effectiveness of the proposed method is as follows. Here we consider the disturbance response design of 3 inertia resonance system using FRIT. The controller here uses the m-IPD (modified Integral-Proportional-Differential) controller for the speed controller. In addition, a PI controller is used for the current controller. FRIT is used to adjust the parameters of the speed controller and the current controller simultaneously. Vibration suppression gains for the step disturbance could be performed to select the property evaluation function.




20

Poster code: E21

Drive of Permanent Magnet Synchronous Motor Using Modular Multi-

Level Inverter

A. Yuto, N. Yoshitake, T. Hanamoto Department of Biological Functions and Engineering, Graduate School of Life Science and

Systems Engineering Kyusyu Institute of Technology, 2-4, Hibbikino, Wakamatsu, Kitakyushu, 808-0196, Japan


Abstract. In recent years, inverters are used in various applications in order to save energy. The inverter is a power conversion equipment that controls the rotation speed of the motor by changing the supply frequency of the motor, and further improvement in efficiency, reduction of noise, and reduction in size is required. Multi-level inverters are one of the type of the inverters, and it has attractive features such as lower harmonics and rate voltage of power semiconductor devices. There is a possibility that a compact and highly efficient inverter can be constructed since it is possible to select small rated and power, low voltage and high speed switching device. Modular multi-level inverter is a series connected full bridge single phase inverter. Increasing the number of the source voltage, the waveform is close to the sinusoidal waveform and the loss of the device is increased, so it is necessary to think about how to construct and use the inverter for various applications. Also, since the magnitude of harmonics and the magnitude of switching loss are different depending on the switching pattern. In this study, the switching methods of level-shift and phase-shift, hysteresis current control are compared, and considered the characteristics of the modular multi-level inverter.




21

Poster code: E22

Calculation Method of Cosmic Ray Failure Rate for Power

Semiconductor Devices in Satellite Electrical Power System

Masaki Sudo, Erdenebaatar Dashdondog, Yuji Shiba, Ichiro Omura Department of Electrical Engineering and Electronics, Kyushu Institute of Technology, 1-

1 Sensui-cho, Tobata-ku, Kitakyushu, Fukuoka, 804-8550, Japan


Abstract. Power semiconductor devices are damaged by localized breakdown due to cosmic ray incidence during high reverse-bias voltage (Fig.1). On the other hand, power consumption in satellites has been increasing in order to large size of satellites. Power generation of a large platform such as space station, big satellites will soon reach the level of MW [1, 2] (Fig.2). This increase brings higher rated voltage of power devices to reduce wire harness mass. Risk of shutting down power supply has been increasing because the failure rate depends on rated voltage. This paper attempts to establish the failure rate calculation method for the sake of improved reliability of power devices in space. We proposed formula of the failure rate by analyzing the failure mechanism, which is applicable to devices on the ground as well as those on satellite orbits. 3.3kV PiN diode model chose to be applied proposed formula was calculated the failure rate. From the result the applied voltage 1700V on the ground got the failure rate 1FIT, whereas the applied voltage 1500V on satellite orbits got the failure rate 1FIT. We can see that the failure rate on satellite orbits is apparently higher than it on the ground (Fig.3). The study concludes that the failure rate contributes to the design of high voltage device for space application.

Fig. 1 Satellite power trend Fig. 2 Satellite power trend Fig. 3 Failure rate VS applied voltage

References: [1] E.E. Haase, C.B. Christensen, H. Ten, Acta Astronautica, vol. 50, No. 12, 2002pp. 747–757. [2] J.R. Brophy, R. Gershman, N. Strange, D. Landau, 47th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit, 2011AIAA, pp. 2011–5514.





22

Poster code: E23

Development of Short Circuit Protection System with PCB Current

Sensor for IGBT Module

M.Ichiki1, S.Tabata1, M.Tsukuda2,3, I.Omura2

1Electrical & Electronic Engineering, Graduate School of Engineering, Kyushu Institute of Technology, 1-1 Sensui-cho, Tobata-ku, Kitakyushu, Fukuoka, 804-8550, Japan

2Department of Biological Functions Engineering, Graduate School of Life Science and Systems Engineering, Kyushu Institute of Technology, 2-4 Hibikino, Wakamatsu-ku,

Kitakyushu, Fukuoka, 808-0196, Japan 3Green Electronics Research Institute, 1-8 Hibikino, Wakamatsu-ku, Kitakyushu, Fukuoka,

808-0135, Japan


Introduction. As the operation power increases and miniaturization of the IGBT development progress, the temperature rise of the chip at short circuit becomes sharp. So, further improvement in the protection speed is required. In this paper, we propose high speed short circuit (SC) protection system for IGBT module. The proposed system Fig.1 is novel SC protection system. Main feature of this system is PCB current sensor and digital circuit. This current sensor made by printing the Rogowski coil on a print substrate and can be integrated in IGBT module due to its compact size (Fig. 2). The digital circuit detects short circuit from the output of the sensor and controls an IGBT.

Result. Fig.3 shows circuit diagram of SC experiment. As a result, the time it takes to protect from short circuit is 660ns in this system (Fig.4). This result shows that the protection speed of this system is faster than in conventional systems.

Conclusions We have constructed a high speed short circuit protection system that can be integrated in an IGBT module using the PCB current sensor.This short circuit protection system is compatible high frequency. The time it takes to protect from short circuit is 660ns in this system. This result shows that the protection speed of this system is faster than in conventional systems.





23

Fig.1 Diagram of novel shrot circuit protection sysem Fig. 2 PCB current sensor and IGBT module

Fig. 3 Circuit diagram of short circuit experiment Fig. 4 (a) Result of SC protection (b) Enlarged view of (a)

[1] L.Dalesandro, N.Karrer, M.Ciappa, A.Castellazzi, W.Fichtner, Power Electronics Specialists Conference, 2008. PESC 2008. IEEE, Pages 2600-2606

[2] M.Koga,M.Tsukuda,K.Nakashima, and I.Omura. Proc. Int. Conf. Integer. Power Electron.Syst.,Mar.2016

[3] E.Motto, J.Donlon, S. Ming, K. Kuriaki, T. Iwagami, H. Kawafuji and T. Nakano. Proc. of IAS’08, pp.1-5 (2008)

[4] M. Kudoh, Y. Hohi, S. Momota, T. Fujiwara and K. Sakurai. Proc. of ISPSD’ 96, pp.303-306 (1996)




24

Poster code: E24

Active Gate Driver to Reduce VCE Overshoot and Turn-off Loss of IGBT

T. Arimoto1, M. Tsukuda2, 3, I. Omura2 1Electrical & Electronic Engineering, Graduate School of Engineering, Kyushu Institute of

Technology, 1-1 Sensui-cho, Tobata-ku, Kitakyushu, Fukuoka, 804-8550, Japan 2Department of Biological Functions Engineering, Graduate School of Life Science and

Systems Engineering, Kyushu Institute of Technology, 2-4 Hibikino, Wakamatsu-ku, Kitakyushu, Fukuoka, 808-0196, Japan

3Green Electronics Research Institute, 1-8 Hibikino, Wakamatsu-ku, Kitakyushu, Fukuoka, 808-0135, Japan

*Corresponding author’s e-mail:[email protected]

Abstract. Switching loss in power devices diminish power conversion efficiency, and voltage surge cause electric magnetic noise and parasitic turn-on. Most of the IGBT losses occur at device turn-off. It is known that loss and surge generated during switching depend on gate resistance. In the conventional gate driver, the relation of trade-off occurs between loss and surge due to use a constant gate resistance. Therefore, we tried to improve the trade-off relationship by using active gate drive which is changing the gate resistance during switching of IGBT. The schematic diagram of the implemented active gate driver is shown in Fig. 1. In order to implement active gate drive, 16 parallel transistors are connected to the gate of the IGBT and a 16-bit control signal is applied to specify the number of activated pull-down npn transistors. Active gate driver was controlled by creating binary data on PC and sending control signal from digital waveform generator. Double pulse test was performed and the voltage current waveform at the time of turning off the low-side IGBT was observed with an oscilloscope. In this measurement, double pulse was input to the gate of the IGBT, and the energy loss at the time of turning off the IGBT and the voltage surge were measured. Experiments were conducted using active gate drive and single step gate drive in order to compare the conventional gate driving method and the active gate driving method. In this study, we’ve optimize the trade-off relationship by using active gate drive, and finally succeeded in reducing VCE overshoot by 56% and turn-off loss by 70% (Fig. 2).

Fig. 1 Schematic diagram of active gate driver Fig. 2 Measured energy loss vs. VCE overshoot in turn-off

characteristics at 40-V switching for Si- IGBT

References [1] Koutarou, M., Seiya, A., Masanori, T., Ichiro, O., Keij, W., Makoto, T., and Takayasu, S. IEEE

Applied Power Electronics Conference and Exposition, pp. 1640-1645, March. 2016. [2] Koutarou, M., Makoto, T. and Takayasu, S. IEEE Energy Conversion Congress and

Exposition, September. 2016. [3] Koutaru, M., Ichiro, O., Makoto, T. and Takayasu, S. IEEE Annual Southern Power

Electronics Conference, pp. 1-4, December. 2016





25

Poster code: E25

Sensorless Control Considering Inductance Change of Synchronous

Reluctance Motor

K.Ando, T.Hanamoto Department of Biological Functions and Engineering, Graduate School of Life Science and

Systems Engineering, Kyushu Institute of Technology, 2-4, Hibikino, Wakamatsu, Kitakyushu, 808-0196, Japan


Abstract. This paper presents the sensorless control of synchronous reluctance motors. In recent years, the importance of power electronics has increased for saving energy in not only the industry but also the home appliance, and an efficient motor is desired. If it is possible to increase the efficiency of motors, you can increase the efficiency of all of the machines that use them. Synchronous reluctance motors haves high efficiency and resource saving is possible because it does not use a permanent magnet. However, it has a characteristic that the value of inductance changes due to the current. Therefore, sensorless control is difficult. Here, we consider the sensorless control of the synchronous reluctance motor with the nonlinear observer and the validity is confirmed using PLECS which is simulation software specialized in power electronics. The observer estimates the unknown state from the system parameters, and input and output values. Here in order to estimate the rotor position required for control of synchronous reluctance motor estimates the induced voltage in a virtual. Since it is necessary to make it a discrete system for use in an actual machine, it was made into a discrete system by using the modified Euler method. Since the modified Euler method can use the continuous system equation as it is, it is easy to reflect changes in inductance. As a result, we confirmed that even if the inductance value changes, proposed sensorless control achieved successfully.

Your Paper's Title Starts Here - Universiti Putra Malaysiasaes2017.upm.edu.my/ENG Poster...

Documents

Transcript of Your Paper's Title Starts Here - Universiti Putra Malaysiasaes2017.upm.edu.my/ENG Poster...