WEKO3
アイテム
キューブサットの太陽光パネルにおけるオンボード異常検知のための機械学習手法の研究と提案
https://doi.org/10.18997/00009176
https://doi.org/10.18997/000091763f2dcea6-cc4a-4833-92c4-8c91fe0b01d6
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
kou_k_569.pdf (2.9 MB)
|
|
| アイテムタイプ | 学位論文 = Thesis or Dissertation(1) | |||||||
|---|---|---|---|---|---|---|---|---|
| 公開日 | 2023-04-07 | |||||||
| 資源タイプ | ||||||||
| 資源タイプ識別子 | http://purl.org/coar/resource_type/c_db06 | |||||||
| 資源タイプ | doctoral thesis | |||||||
| タイトル | ||||||||
| タイトル | Study and Proposal of Machine Learning Methods for On-board Anomaly Detection in CubeSats Solar Panels | |||||||
| 言語 | en | |||||||
| タイトル | ||||||||
| タイトル | キューブサットの太陽光パネルにおけるオンボード異常検知のための機械学習手法の研究と提案 | |||||||
| 言語 | ja | |||||||
| 言語 | ||||||||
| 言語 | eng | |||||||
| 著者 |
Jara, Adolfo
× Jara, Adolfo
|
|||||||
| 抄録 | ||||||||
| 内容記述タイプ | Abstract | |||||||
| 内容記述 | The development of small satellites by non-space-faring nations is significantly driven by the availability of low-cost launch and commercial off-the-shelf （COTS） components. The BIRDS program follows the approach of a Lean Satellite concept. The concept relies on utilizing commercially available yet non-space-proven components to obtain effective and efficient development. On the other hand, the mass and size limitations usually indicate only a few or no redundant systems available. The intermittent and short-term communication window limits the data transmission capability, potentially affecting the housekeeping-data-monitoring analysis, the primary way to verify the status of the satellites in operation. Furthermore, the CubeSat system’s limitations generally include power generation, telemetry bandwidth, computational power, and memory. BIRDS-4 is the fourth iteration of the BIRDS program. It is a constellation of three CubeSats: GuaraniSat-1, Maya-2, and Tsuru, deployed into orbit on 14 March 2021. BIRDS-4 team discovered a critical issue related to the electrical power subsystem of GuaraniSat-1. The team could not immediately detect the symptoms and the satellite stopped transmitting after three days of operation. Subsequent analysis of the continuous wave （CW） beacon data confirmed the no power generation on two solar panels. During BIRDS-3 satellites operation, two satellites have revealed power generation loss in one of their panels. Therefore, this research presents a Machine Learning （ML） approach for on-board anomalies detection in CubeSat´s telemetry data of the solar panel system. Five ML algorithm candidates are investigated, considering classification score, execution time, model size, and power consumption in a constrained computational environment. The approach considers both pre-processing methods and ML models. The pre-processing stage introduces the windowed averaging technique besides standardization and principal component analysis. The process is applied to a solar panel’s dataset generated by BIRDS-3 and BIRDS-4 satellites. It is important to note that this is the first utilization of such data. The outcomes will be beneficial as a first step towards machine learning utilization on-board CubeSats that will be a basis to address anomaly detection issues that may arise in space. | |||||||
| 目次 | ||||||||
| 内容記述タイプ | TableOfContents | |||||||
| 内容記述 | 1. Introduction||2. State of Art||3. Materials||4. Methods||5. Results||6. Conclusions And Future Work | |||||||
| 備考 | ||||||||
| 内容記述タイプ | Other | |||||||
| 内容記述 | 九州工業大学博士学位論文 学位記番号: 工博甲第569号 学位授与年月日: 令和5年3月24日 | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Anomaly Detection | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Electrical Power System | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | CubeSat | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | Machine Learning | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | BIRDS project | |||||||
| アドバイザー | ||||||||
| 趙, 孟佑 | ||||||||
| 学位授与番号 | ||||||||
| 学位授与番号 | 甲第569号 | |||||||
| 学位名 | ||||||||
| 学位名 | 博士（工学） | |||||||
| 学位授与年月日 | ||||||||
| 学位授与年月日 | 2023-03-24 | |||||||
| 学位授与機関 | ||||||||
| 学位授与機関識別子Scheme | kakenhi | |||||||
| 学位授与機関識別子 | 17104 | |||||||
| 学位授与機関名 | 九州工業大学 | |||||||
| 学位授与年度 | ||||||||
| 内容記述タイプ | Other | |||||||
| 内容記述 | 令和4年度 | |||||||
| 出版タイプ | ||||||||
| 出版タイプ | VoR | |||||||
| 出版タイプResource | http://purl.org/coar/version/c_970fb48d4fbd8a85 | |||||||
| アクセス権 | ||||||||
| アクセス権 | open access | |||||||
| アクセス権URI | http://purl.org/coar/access_right/c_abf2 | |||||||
| ID登録 | ||||||||
| ID登録 | 10.18997/00009176 | |||||||
| ID登録タイプ | JaLC | |||||||
