Search

Amid the bloom of Renewable energy (RE) integrated into the grid, an accurate Photovoltaic(PV) power forecast is considered to be a crucial task in maintaining the reliability and stability of the power systems since this technology strongly depends on various external factors, causing the fluctuation in the output power. However, the poor quality of input data, which is very common in practical circumstances owing to the low-cost measurement and data acquisition devices, poses an enormous challenge for the predictive model to deeply extract the spatial and temporal correlation of the input data. This study proposes a Multi Two-Dimensional Convolutional Neural Network (2D-CNN) for short-term PV power forecast embedded with Laplacian Attention mechanism. By viewing the input sequences in a 2D form, the input map is constructed, and the interconnected feature among variables can be captured by convolution operation. Moreover, with the multiple CNN layers working in parallel architecture, different representations hidden inside the input map can be detected, enabling the proposed model to bring out promising performance across forecast time-step without modifying its initial parameters. In order to reduce the decay impact of irrelevant variables existing inside the input data, the Laplacian Attention mechanism is employed. The Attention matrix is dynamically modified during the training process to produce an accurate attention matrix, which represents the correlation between variables. Therefore, the model is able to focus on informative features and ignore negative ones. The experiments conducted on two different datasets with opposite characteristics provide deep insights into the strength of the proposed model over the baseline model, which strongly demonstrates the efficiency of the proposed model, especially when dealing with datasets bearing tough characteristics.

In Dual-Active-Bridge converter control system, high frequency terminal current is usually measured then fed-back to the current mode controller. However, the measurement requires wide bandwidth sensors, fast and precise operational amplifier, and high sampling rate Analog to Digital conversion. This paper proposes a Proportional Integral Observer to estimate such the current to eliminate the need of current sensors. A combined feedforward-feedback control system is then developed to enhance the system dynamics and to regulate the output voltage. The control system needs only the information of terminal voltages. Simulation and experiment results show that the observer performs fast and accurately with acceptable observation error; and the output voltage is well regulated regardless of load changes.

AbstractThis paper proposes an observer-based digital control system to manage active and reactive powers in a Dual-Active-Bridge converter. Since they are proportional to the direct and the quadrature components of the transferred current, Idand Iq, respectively, the power management is realized by regulating the two current components. An observer is designed to estimate them from the feedback voltage and the average absolute value of the transferred current. Since all the feedback signals for the observer are DC quantities, the sampling speed is not necessary to be very high. A digital controller is then synthesized to regulate Idand Iqat desired values. In the control system, the bridge shift angle is used to handle Iq, meanwhile the switching frequency is variable to manipulate Id. The current command for Idis calculated from the desired active power, meanwhile, by assigning a positive value for Iqcontrol loop, soft turn on is easily achieved. The performance of the observer as well as the digital control system is validated by a laboratory-scale experiment system.

AbstractThis paper presents a module approach for lab-scale wind power systems (WPS). The system is designed into functional modules. These modules can be used independently or in other projects. It is also convenient to handle and set up them on a small table in classrooms and labs. Hence, they are very useful for training and research. In addition, operation on both aerodynamic side and generator side is developed in this research. Many papers on wind system simulators focus only on aerodynamic system with pitch angle control or generator system with optimal control. Thus, the operation of WPS in 2 regions, which requires coordination of aerodynamic and generator sides, is not explained and demonstrated in previous study. Many case studies are conducted to demonstrate the capability of the module-based WPS. The system is capable of speed control, optimal power control and pitch angle control. As illustrated in experimental results, these control strategies can efficiently cooperate under operation in 2 regions of WPS.

This paper presents an application of Cuckoo search algorithm to determine optimal location and sizing of Static VAR Compensator. Cuckoo search algorithm is a modern heuristic technique basing Cuckoo species’ parasitic strategy. The Lévy flight has been employed to generate random Cuckoo eggs. Moreover, the objective function is a multiobjective problem, which minimizes loss power, voltage deviation and investment cost of Static VAR Compensator while satisfying other operating constraints in power system. Cuckoo search algorithm is evaluated on three case studies and compared with the Teaching-learning-based optimization, Particle Swarm optimization and Improved Harmony search algorithm. The results show that Cuckoo search algorithm is better than other optimization techniques and its performance is also better.

AbstractCurrently in Japan, activities for the reduction of nuclear energy have increased in the wake of Great East Japan Earthquake. So smart-grid and micro-grid incorporating renewable energies such as photovoltaic power, small hydroelectric generation and wind power are paid attention to. In this research, the power supply system is built for an area suitable for a condition that there is the least of the energy demand and much renewable energy quantity of natural resource. This time, we assume the introduction of small hydropower generator in this energy system. So it is necessary to consider the effect of the device in the cell. We clarify the following discussion items by instantaneous analysis. First, the capacities of the inverter power supply and induction generator for synchronous generator are increased. Second, evaluation of the system reliability and stability by frequency variation and voltage fluctuations is presented. Finally, in conclusion, the optimal capacities of the induction generator and inverter power supply are determined.

AbstractThis research proposes a sensorless Maximum Power Point Tracking (MPPT) method for a hybrid Photovoltaic-Wind system, which consists of Photovoltaic (PV) system and Doubly-Fed Induction Generator (DFIG) Wind Turbine. In the hybrid system, the DC/DC converter output of the PV system is directly connected to the DC-link of DFIG’s back-to-back converter. Therefore, the PV inverter and its associated circuit can be removed in this structure. Typically, the PV power is monitored by using PV current sensor and PV voltage sensor for MPPT. In this paper, the powers of converters on grid side and rotor side of DFIG are used to estimate the PV power without the PV current sensor. That can efficiently reduce the cost of the hybrid system. The detailed analysis of the sensorless MPPT method, which includes derived equations and operation response, is also presented in this paper. In addition, an overview of PV-DFIG research in literature is stated to supply comprehensive knowledge of related research.

About half of northeastern Asians lack ALDH2 (Acetaldehyde Dehydrogenase 2), an enzyme involved in alcohol metabolism. People with deficient ALDH2 often experience facial flushing after drinking a small dose of alcohol. The present study examined the effect of a small dose of alcohol on driving-related behavior of northeastern Asians (Japanese). Participants were 9 Japanese flushers and 12 Japanese nonflushers. They were evaluated using a driving simulator and self-report of intoxication after ingesting 0.6g/kg or 0.4g/kg of alcohol or a placebo. Facial flushers were more likely to experience subjective feelings of intoxication than were nonflushers. There were no significant differences between facial flushers and nonflushers in blood alcohol concentration and choice reaction time to visual stimuli. Despite having different subjective feelings of intoxication, facial flushers and nonflushers were equally impaired after having ingested equal amounts of alcohol per body weight.

This paper proposes a new load-shedding method which can be implemented locally. A sudden outage of generation system needs quick and appropriate load-shedding, which must satisfy some conditions to make the whole system stable. The proposed method is based on the use of power flow deviations observed at each local bus just after the disturbance, which make possible autonomous controls. In order to exemplify the effectiveness of the proposed method, two suppressing evaluations, voltage deviations and line overload are discUssed. The result is that the proposed method controls almost the same load-shedding distribution by full optimizing method, so that it can implement improved load-shedding as compared with the conventional method.

This paper describes an estimation method of the amount of power drop after severe generation outage. Power demand always must be equal to power supply for keeping system frequency in criterion value. So once a generation outage accident occurs, adequate value of loads must be shed within a few seconds. It is difficult to estimate accurate amount of generation outage from this frequency. The proposed method is based on an estimation method of system's dominating differential equation for analyzing inverse problems. New method is induced by adding supplementary information which makes up for lacked information. Model simulations indicate that the proposed method provides good estimation comparing with a conventional one.

We demonstrated volumetric optical recording with void marks in a bulk recording medium using a newly developed all-semiconductor picosecond laser. The laser is compact and has an ultrahigh peak power, which is necessary for practical application of the volumetric optical recording system. We confirmed the laser to be able to record void marks effectively on our recording media. The laser was implemented to our optical drive system as a light source and as much as 30 layer void-mark recording was carried out on the 200-μm-thick bulk recording medium. The total recording capacity was 97 Gbyte and the bit error rates (bERs) were measured to be on the order of $10^{-4}$ for most of the layers. We believe that this was a significant step to realize a practical volumetric optical recording system.

We have demonstrated volumetric optical recording using an all-semiconductor picosecond laser, which generated optical pulses with a duration of 3 ps and a maximum peak power of 100 W at a wavelength of 404 nm and a repetition frequency of 1 GHz. This pulsed laser system efficiently induced multiphoton absorption in the recording media due to its high peak power and high repetition rate. The recording marks were formed as submicrometer voids inside a single thick recording layer by multiphoton absorption. A clear readout signal was obtained from the recorded marks.

A new magneto-optical recording method of magnetic field modulation with pulsed laser irradiation has been developed. In this paper, we report the theory of the recording process and the characteristics of the method. This method is appropriate for high power-efficiency and high density edge-recording.

We have studied the feasibility of the 120-mm-sized magneto optical disk system which achieves 6-GB user capacity using an optical pickup with a 640-nm wavelength laser and a 0.6-numerical aperture of the objective lens. We have used three technologies: a laser-pulsed magnetic-field modulation recording, a center-aperture-detection type of magnetically-induced super resolution disk with a 0.6-mm-thick single substrate and partial response equalization with maximum likelihood detection. We have evaluated the system-margins of write power, radial tilt, tracking and focusing offset and have confirmed the feasibility of the linear bit density of 0.235 µm/bit in 0.6-µm track pitches.

A new structure disk was developed for a magneto-optical (MO) composite head system. The new disk has a thin readout substrate under and a high magnetic permeability layer over the MO layer. The effect of the high magnetic permeability layer was confirmed experimentally and agreed well with theoretical results. The normal component of the magnetic field at the MO layer is enhanced 1.6 times by the high magnetic permeability layer.