WEKO3
アイテム
水産養殖における給餌支援を目的とした養殖魚活動量推定センサネットワーク
https://doi.org/10.18997/00009082
https://doi.org/10.18997/000090821e0af7e0-b888-4a60-9b44-4345abb5adc3
| 名前 / ファイル | ライセンス | アクション |
|---|---|---|
sei_k_447.pdf (11.3 MB)
|
|
| アイテムタイプ | 学位論文 = Thesis or Dissertation(1) | |||||||
|---|---|---|---|---|---|---|---|---|
| 公開日 | 2023-02-08 | |||||||
| 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 | A network of flow speed sensors estimating fish activity for assisting in feeding decisions in cage aquaculture | |||||||
| 言語 | en | |||||||
| タイトル | ||||||||
| タイトル | 水産養殖における給餌支援を目的とした養殖魚活動量推定センサネットワーク | |||||||
| 言語 | ja | |||||||
| 言語 | ||||||||
| 言語 | eng | |||||||
| 著者 |
Bautista, Solpico Dominic
× Bautista, Solpico Dominic
|
|||||||
| 抄録 | ||||||||
| 内容記述タイプ | Abstract | |||||||
| 内容記述 | With the expansion of aquaculture production to meet the growing demand for food fish worldwide, there is an increasing need for its sustainable management not only to mitigate any threat to the aquatic environment but also to produce more high-quality fish that meet the market standards for seafood. Digital transformation （DX） holds an important role in achieving this need, enabling fish farmers make better decisions in using their resources as well as in reducing their costs of production through knowledge transfer and data. One such decision-making where DX can assist is in feeding, which generally has the largest share in production costs. Conventionally, farmers control the feeding from judging the fishes’ behavior. They learn this practice through their subjective experiences, leading to substantial differences in results between expert and novice farmers. The latter tends to feed the fishes inefficiently, producing uneaten feeds, which do not only increase financial burden in culture operations but also contribute to the pollution of the aquatic environment, which affect the growth and quality of the fish stocks and ultimately the sustainability of their operations. Applying DX to estimate the fish behavior therefore becomes important. While several intelligent feeding control methods using various technologies have been developed for applying such DX, many of are either easily affected by changes or noise from external sources or are technically difficult to implement in larger scales. An alternative approach is by measurement of outward flow from the cage, which has been observed to be fish induced. If we assume that fishes tend to swim upward when they sense feeds coming from the surface and swim back down when satiated, and that they tend to move in circles, fish activity at different depths can be visualized with this measurement to help farmers make feeding decisions. An off-grid modular sensor network was thus designed and developed to collect flow speed measurements and underwater video recordings from at least two depths and from multiple sides of a fish cage. This was realized by organizing sensors into sensor modules, which are connected to each other and are organized into sensor units. The sensor units were designed wirelessly relay data from all modules to a hub unit. Flow sensors were modified to measure flow speeds underwater. The network’s operation was also designed to be scheduled to manage its offshore power supply to enable long-term observation by the system. To demonstrate its functionality, the sensor network was deployed in fish cages and collected data, especially during feeding. The flow measurements and underwater videos were analyzed together to estimate the fish activity. Although there were various patterns, it could be observed that surface flow increased significantly at the beginning of feeding and declined toward the end. Vigorous surface activity was observed at most cages, validating the observed flow speeds. Offset between speeds at opposite sides was also observed, suggesting cancellation of global currents. In some experiments, increase of flow below the surface was also observed at the beginning and towards the end of feeding, indicating fishes to climbing and descending. There are many factors that contribute to the speed of flow coming out of the cage. However, the fishes’ locomotion and depth distribution have a large contribution to the changes in flow speed. These parameters depend on their hunger level and on the availability of feeds in water, as fishes may tend to swim up fast when they sense feed in water and swim less vigorously when they start to become full. A simplified model of the fish activity as a response to feeding was developed for simulation. Its output could be compared with the collected flow data for the farmers to use in improving their feeding decisions. Some observed flow patterns such as the decline of surface flow and the increase of flow at lower depths could be used for deciding on when to stop feeding. With these insights, a DX system was envisioned to collect flow speed and other measurements from multiple fish cages, assisting fish farmers in feeding. This research contributes to the development of DX application in cage aquaculture by introducing a flexible self-correcting system that could help farmers visualize underwater fish activity to help them improve their feeding decisions. | |||||||
| 言語 | en | |||||||
| 目次 | ||||||||
| 内容記述タイプ | TableOfContents | |||||||
| 内容記述 | 1. Introduction||2. Flow speed sensor network||3. Fish activity estimation||4. Discussion on feeding decision||5. Conclusion | |||||||
| 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：生工博甲第447号 学位授与年月日：令和4年9月26日 | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | cage aquaculture | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | feeding behavior | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | sensor network | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | flow speed | |||||||
| キーワード | ||||||||
| 主題Scheme | Other | |||||||
| 主題 | fish activity | |||||||
| アドバイザー | ||||||||
| 石井, 和男 | ||||||||
| 学位授与番号 | ||||||||
| 学位授与番号 | 甲第447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 | 2022-09-26 | |||||||
| 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 | 10.18997/00009082 | |||||||
| ID登録タイプ | JaLC | |||||||
