Digital Library[ Search Result ]
A Diffusion-based Trajectory Prediction Model for Flight Vehicles Considering Pull-up Maneuvers
Seonggyun Lee, Joonseong Kang, Jeyoon Yeom, Dongwg Hong, Youngmin Kim, Kyungwoo Song
http://doi.org/10.5626/JOK.2025.52.3.241
This paper proposes a new model for processing multivariate time series data aimed at predicting nonlinear trajectories related to aircraft pull-up maneuvers. To achieve this, aircraft trajectories were predicted using CSDI (Conditional Score-based Diffusion Models for Imputation), a state-of-the-art generative AI model. Specifically, because the flight distance and shape of the aircraft vary significantly depending on the presence of pull-up maneuvers, the data were separated into subsets with and without these maneuvers to train and predict distinct models. Experimental results demonstrated that the model predicted trajectories very similar to actual trajectories and achieved superior performance in MAE, RMSE, and CRPS metrics compared to existing deep learning models. This study not only enhances the accuracy of aircraft trajectory prediction but also suggests the potential for more sophisticated predictions through future integration with Classifier Diffusion models.
Photovoltaic Power Forecasting Scheme Based on Graph Neural Networks through Long- and Short-Term Time Pattern Learning
Jaeseung Lee, Sungwoo Park, Jaeuk Moon, Eenjun Hwang
http://doi.org/10.5626/JOK.2024.51.8.690
As the use of solar energy has become increasingly common in recent years, there has been active research in predicting the amount of photovoltaic power generation to improve the efficiency of solar energy. In this context, photovoltaic power forecasting models based on graph neural networks have been presented, going beyond existing deep learning models. These models enhance prediction accuracy by learning the interactions between regions. Specifically, they consider how the amount of photovoltaic power in a specific region is affected by the climate conditions of adjacent regions and the time pattern of photovoltaic power generation. However, existing models mainly rely on a fixed graph structure, making it difficult to capture temporal and spatial interactions. In this paper, we propose a graph neural networks-based photovoltaic power forecasting scheme that takes into account both long-term and short-term time patterns of regional photovoltaic power generation data. We then incorporate these patterns into the learning process to establish correlations between regions. Compared to other graph neural networks-based prediction models, our proposed scheme achieved a performance improvement of up to 7.49% based on the RRSE, demonstrating its superiority.
A GRU-based Time-Series Forecasting Method using Patching
http://doi.org/10.5626/JOK.2024.51.7.663
Time series forecasting plays a crucial role in decision-making within various fields. Two recent approaches, namely, the patch time series Transformer (PatchTST) and the long-term time series foraging linear (LTSF-Linear) of the MLP structure have shown promising performance in this area. However, PatchTST requires significant time for both model training and inference, while LTSF-Linear has limited capacity due to its simplistic structure. To address these limitations, we propose a new approach called patch time series GRU (PatchTSG). By leveraging a Gated Recurrent Unit (GRU) on the patched data, PatchTSG reduces the training time and captures valuable information from the time series data. Compared to PatchTST, PatchTSG achieves an impressive reduction in learning time (up to 82%) and inference time (up to 46%).
Time-Series Data Augmentation Based on Adversarial Training
http://doi.org/10.5626/JOK.2023.50.8.671
Recently, time series data are being generated in various industries with advancement of the Internet of Things (IoT). Accordingly, demands for time series forecasting in various industries are increasing. With acquisition of a large amount of time-series data, studies on traditional statistical method based time-series forecasting and deep learning-based forecasting methods have become active and the need for data augmentation techniques has emerged. In this paper, we proposed a novel data augmentation method for time series forecasting based on adversarial training. Unlike conventional adversarial training, the proposed method could fix the hyperparameter about the number of adversarial training iterations and utilize blockwise clipping of perturbations. We carried out various experiments to verify the performance of the proposed method. As a result, we were able to confirm that the proposed method had consistent performance improvement effect on various datasets. In addition, unlike conventional adversarial training, the necessity of blockwise clipping and the hyperparameter value fixing proposed in this paper were also verified through comparative experiments.
Early Anomaly Detection of LNG-Carrier Main Engine System based on Multivariate Time-Series Boundary Forecasting and Confidence Evaluation Technique
Donghyun Kim, Taigon Kim, Minji An, Yunju Baek
http://doi.org/10.5626/JOK.2023.50.5.429
Recently, a variety of studies have been conducted to detect abnormal operation of ships and their causes and in the marine and shipbuilding industries. This study proposed a method for early anomaly detection of the main engine system using a multivariate time series sensor data extracted from LNG carriers built at a shipyard. For early anomaly detection, the process of predicting the future value through the sensor data at present is necessary, and in this process, the prediction residual, which is the difference between the actual future value and the predicted value, is generated. Since the generated residual has a significant effect on the early anomaly detection results, a compensating process is necessary. We propose novel loss functions that can learn the upper or lower prediction boundary of a time-series forecasting model. The time-series forecasting model trained with the proposed loss function improves the performance of the early anomaly detection algorithm by compensating the prediction residual. In addition, the real-time confidence of the predicted value is evaluated through the newly proposed confidence model by utilizing the similarity between time-series forecasting residual and confidence residual. With the early anomaly detection algorithm proposed in this study, the prediction model, which learns the upper boundary, outputs the upper limit of the predicted value that can be output by the baseline prediction model learned with the MSE loss function and can predict abnormal behavior that threshold-based anomaly discriminator could not predict because the future prediction of the baseline model is lower than the actual future value. Based on the results of this study, the performance of the proposed method was improved to 0.9532 compared to 0.4001 of the baseline model in Recall. This means that robust early anomaly detection is possible in various operating styles of the actual ship operations.
Denoising Multivariate Time Series Modeling for Multi-step Time Series Prediction
Jungsoo Hong, Jinuk Park, Jieun Lee, Kyeonghun Kim, Seung-Kyun Hong, Sanghyun Park
http://doi.org/10.5626/JOK.2021.48.8.892
The research field of time series forecasting predicts the future time point using seasonality in time series. In the industrial environment, since decision-making through continuous perspective prediction of the future is important, multi-step time series forecasting is necessary. However, multi-step prediction is highly unstable because of its dependency on predicted value of previous time prediction result. Therefore, the traditional time series forecasting makes a statistical prediction for the single time point. To address this limitation, we propose a novel encoder-decoder based neural network named ‘DTSNet’ which predicts multi-step time points for multivariate time series. To stabilize multi-step prediction, we exploit positional encoding to enhance representation for time point and propose a novel denoising training method. Moreover, we propose dual attention to resolve long-term dependencies and modeling complex patterns in time series, and we adopt multi-head strategy at linear projection layer for variable-specific modeling. To verify the performance improvement of our approach, we compare and analyze it with baseline models, and we demonstrate the proposed methods through comparison tests, such as, component ablation study and denoising degree experiment.
A Feature Selection Technique in the Neural Network for Demand Forecasting of Mobile Payment System
Ho-Joon Kim, Yun-Seok Cho, Kyungmi Kim
http://doi.org/10.5626/JOK.2018.45.4.370
In this paper, we present a time series prediction technique based on neural network as a methodology for forecasting service demand of mobile payment system. We propose a two-stage neural network model for the feature selection process and the prediction process. Three types of fuzzy membership functions were adopted for the representation of feature data, and a hyperbox-based neural network model is used for the evaluation of feature relevance factor. The proposed feature selection technique reduces the amount of computation and eliminates erroneous feature data in the learning data set. We evaluated the usefulness of the proposed method through experiments using two years of data obtained form actual smart campus systems.
‘Hot Search Keyword’ Rank-Change Prediction
Dohyeong Kim, Byeong Ho Kang, Sungyoung Lee
http://doi.org/10.5626/JOK.2017.44.8.782
The service, "Hot Search Keywords", provides a list of the most hot search terms of different web services such as Naver or Daum. The service, bases the changes in rank of a specific search keyword on changes in its users’ interest. This paper introduces a temporal modelling framework for predicting the rank change of hot search keywords using past rank data and machine learning. Past rank data shows that more than 70% of hot search keywords tend to disappear and reappear later. The authors processed missing rank value, using deletion, dummy variables, mean substitution, and expectation maximization. It is however crucial to calculate the optimal window size of the past rank data. We proposed an optimal window size selection approach based on the minimum amount of time a topic within the same or a differing context disappeared. The experiments were conducted with four different machine-learning techniques using the Naver, Daum, and Nate "Hot Search Keywords" datasets, which were collected for 2 years.
Search
Journal of KIISE
- ISSN : 2383-630X(Print)
- ISSN : 2383-6296(Electronic)
- KCI Accredited Journal
Editorial Office
- Tel. +82-2-588-9240
- Fax. +82-2-521-1352
- E-mail. chwoo@kiise.or.kr