• Korean Journal of Remote Sensing
• /
• v.39 no.5_3
• /
• pp.979-995
• /
• 2023

As wildfires are difficult to predict, real-time monitoring is crucial for a timely response. Geostationary satellite images are very useful for active fire detection because they can monitor a vast area with high temporal resolution (e.g., 2 min). Existing satellite-based active fire detection algorithms detect thermal outliers using threshold values based on the statistical analysis of brightness temperature. However, the difficulty in establishing suitable thresholds for such threshold-based methods hinders their ability to detect fires with low intensity and achieve generalized performance. In light of these challenges, machine learning has emerged as a potential-solution. Until now, relatively simple techniques such as random forest, Vanilla convolutional neural network (CNN), and U-net have been applied for active fire detection. Therefore, this study proposed an active fire detection algorithm using state-of-the-art (SOTA) deep learning techniques using data from the Advanced Himawari Imager and evaluated it over East Asia and Australia. The SOTA model was developed by applying EfficientNet and lion optimizer, and the results were compared with the model using the Vanilla CNN structure. EfficientNet outperformed CNN with F1-scores of 0.88 and 0.83 in East Asia and Australia, respectively. The performance was better after using weighted loss, equal sampling, and image augmentation techniques to fix data imbalance issues compared to before the techniques were used, resulting in F1-scores of 0.92 in East Asia and 0.84 in Australia. It is anticipated that timely responses facilitated by the SOTA deep learning-based approach for active fire detection will effectively mitigate the damage caused by wildfires.

• Journal of Power Electronics
• /
• v.8 no.1
• /
• pp.91-100
• /
• 2008

The d-q harmonic detecting algorithms are dominant methods to generate current references for active power filters (APF). They are often implemented in the synchronous frame and time domain. This paper researches the frequency characteristics of d-q synchronous transformations, which are closely related to the analysis and design issues of control system. Intuitively, the synchronous transformation is explained with amplitude modulation (AM) in this paper. Then, the synchronous filter is proven to be a time-invariant and linear system, and its transfer function matrix is derived in the stationary frames. These frequency-domain models imply that the synchronous transformation has an equivalent effect of frequency transformation. It is because of this feature, the d-q method achieves band-pass characteristics with the low pass filters in the synchronous frame at run time. To simplify these analytical models, an instantaneous positive-negative sequence frame is proposed as expansion of traditional symmetrical components theory. Furthermore, the synchronous filter is compared with the traditional bind-pass filters based on these frequency-domain analytical models. The d-q harmonic detection methods are also improved to eliminate the inherent coupling effect of synchronous transformation. Typical examples are given to verify previous analysis and comparison. Simulation and experimental results are also provided for verification.

• Proceedings of the KIEE Conference
• /
• 2006.07a
• /
• pp.220-221
• /
• 2006

This paper presents a single-phase parallel active power filter with an analog control circuit to eliminate for harmonic source currents generated by nonlinear loads. The proposed system removes the harmonic source currents by injecting a compensation current that is 180' out of phase with the load harmonic current. The detection of the load harmonics is realized by a simple new structure, referred to the Notch Filter with GIC (Generalized Impedance Converter), which has higher Q than existing harmonic detecters and a simpler structure. The compensation current is obtained using the proposed harmonic detection circuit, DC-Link voltage, and output current of the full-bridge inverter controlled current mode PWM controller. The operation of the proposed system is verified experimentally.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.10 no.3
• /
• pp.296-301
• /
• 2005

This paper proposing the use of the Instantaneous reactive power method as a harmonic detection method for a single phase active filter system. This method is to detect harmonic components through d-q frame approach. The conventional use of d-q frame approach for a 3-phase system Is extended to the single phase system. The proposed system uses a multi-level inverter for harmonic compensation and the inverter is connected to the input side without using transformers. The proposed algorithm is verified by simulation and experiment.

• Journal of the Korean Society of Safety
• /
• v.6 no.4
• /
• pp.13-20
• /
• 1991

Ion-selective electrode responsive to the thiocynate ion prepared by using the quaternary ammonium salts as a active material and PVC as a membrane matrix. The effect of chemical structure and composition of active material, and the membrane thickness on the linear response. the detection limit, and Nernstian slope of the electrode studied. Under the above optimum conditions of membrane, the effect of pH and the selectivity coefficients to various interfering anions were compared and investigated. It was concluded that the functions of thiocynate ion-selective electrode(ISE) were closely related to the chemical structure of the quaternary ammonium salts. The linear response, and the detection limit of the electrode potential increased with the increase of the carbon chain length of the alkyl group in the quaternary ammonium salts in the ascending order of Aliquat 336T, TOAT, TDAT, and TDDAT. The optimum membrane thickness was 0.3mm. The electrode characteristics was better with the decrease of the concentration of active material, and the best concentration was 3 weight percent. The membrane potential was independent of the pH variation in the region from pH 2 to 12. The order of the selectivity coefficients is as follows：Cl $O_4$$^{[-10]}$ ＞ $I^{[-10]}$ ＞N $O_3$$^{[-10]}$ ＞B $r^{[-10]}$ ＞ $F^{[-10]}$ ＞C $l^{[-10]}$ ＞O $A_{c}$ $^{[-10]}$ 〓S $O_4$$^{2-}$.

• The KIPS Transactions:PartC
• /
• v.11C no.1
• /
• pp.89-98
• /
• 2004

AQM (Active Queue Management) techniques such as RED (Random Early Detection) which be proposed to solve the congestion of internet perform congestion control effectively for TCP data. However, in the situation where TCP and UDP share the bottleneck link, they can not solve the problems of the unfairness and long queueing delay. In this paper, we proposed an simple queue management algorithm, called PSRED (Protocol Sensitive RED), that improves fairness and decreases queueing delay. PSRED algorithm improves fairness and decreases average queue length by distinguishes each type of flow in using protocol field of packets and applies different drop functions to them respectively.

• Journal of Communications and Networks
• /
• v.5 no.3
• /
• pp.275-281
• /
• 2003

Random Early Detection (RED) [1] is an active queue management scheme which has been deployed extensively to reduce packet loss during congestion. Although RED can improve loss rates, its performance depends severely on the tuning of its operating parameters. The idea of adaptively varying RED parameters to suit the network conditions has been investigated in [2], where the maximum packet dropping probability $max_p$ has been varied. This paper focuses on adaptively varying the queue weight $\omega_q$ in conjunction with $max_p$ to improve the performance. We propose two algorithms viz., $\omega_q$-thresh and $\omega_q$-ewma to adaptively vary $\omega_q$. The performance is measured in terms of the packet loss percentage, link utilization and stability of the instantaneous queue length. We demonstrate that varying $\omega_q$ and $max_p$ together results in an overall improvement in loss percentage and queue stability, while maintaining the same link utilization. We also show that $max_p$ has a greater influence on loss percentage and queue stability as compared to $\omega_q$, and that varying $\omega_q$ has a positive influence on link utilization.

• Journal of Communications and Networks
• /
• v.14 no.1
• /
• pp.34-39
• /
• 2012

Collisions between tags greatly reduce the identification speed in radio frequency identification (RFID) systems and increase communication overhead. In particular for an active RFID system, tags are powered by small batteries, and a large number of re-transmissions caused by collisions can deteriorate and exhaust the tag energy which may result in missing tags. An efficient collision-free arbitration protocol for active RFID systems is proposed in this paper. In this protocol, a new mechanism involving collision detection, collision avoidance, and fast tag access is introduced. Specifically, the pulse burst duration and busy-tone-detection delay are introduced between the preamble and data portion of a tag-to-reader (T-R) frame. The reader identifies tag collision by detecting pulses and transmits a busy tone to avoid unnecessary transmission when collision occurs. A polling process is then designed to quickly access the collided tags. It is shown that the use of the proposed protocol results in a system throughput of 0.612, which is an obvious improvement when compared to the framed-slotted ALOHA (FSA) arbitration protocol for ISO/IEC 18000-7 standard. Furthermore, the proposed protocol greatly reduces communication overhead, which leads to energy conservation.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.36 no.6
• /
• pp.550-557
• /
• 2008

Delamination damage at a bolt hole in a composite stiffened panel was detected using a built-in piezoelectric active sensor array. Various signal processing techniques were used to detect an invisible small scale delamination around a fastener hole due to localized transverse loading. A built-in piezoelectric sensor array was used to generate diagnostic signals and to measure response signals. Then, the response signals were processed to extract damage-sensitive features. Damage indexes were calculated to estimate the severity and location of the damage from the features.

• Smart Media Journal
• /
• v.6 no.3
• /
• pp.49-56
• /
• 2017

Images and Videos that include the human face contain a lot of information. Therefore, accurately extracting human face is a very important issue in the field of computer vision. However, in real life, human faces have various shapes and textures. To adapt to these variations, A model-based approach is one of the best ways in which unknown data can be represented by the model in which it is built. However, the model-based approach has its weaknesses when the motion between two frames is big, it can be either a sudden change of pose or moving with fast speed. In this paper, we propose an enhanced human face-tracking model. This approach included human face detection and motion estimation using Cascaded Convolutional Neural Networks, and continuous human face tracking and modeling correction steps using the Active Appearance Model. A proposed system detects human face in the first input frame and initializes the models. On later frames, Cascaded CNN face detection is used to estimate the target motion such as location or pose before applying the old model and fit new target.