발간년도 : [2023]
| 논문정보 |
|
| 논문명(한글) |
[Vol.18, No.6] Research on the Development of Smart Farm Sensor Status-based Prognostics and Health Management Algorithms |
|
| 논문투고자 |
Hyeon-O Choe, Myeong-Hoon Lee |
|
| 논문내용 |
Recently, the direction of agriculture in Korea is aimed at realizing highly efficient digital agriculturethrough science and technology. Accordingly, digital agricultural technology is spreading nationwide, butthe reliability and accuracy of sensors and data are not secured, reducing the usability of smart farmsand the effectiveness of policies. Therefore, in this paper, we aim to apply Prognostics and HealthManagement (PHM) technology to develop sensor failure symptom detection and diagnosis technologythrough analysis of sensor environment data used in smart farms. Five farms were selected to collectsmart farm data, and humidity sensors and CO₂ sensors, which were causing problems in smart farmgreenhouses, were selected as PHM management targets. Using health management technology, abnormaldata detection and failure prediction technology was developed using a mathematical model and amachine learning-based abnormality detection algorithm. As a result of the experiment, the relativehumidity estimation algorithm identified more than 90% of the abnormal data and sensor failureconditions and responded to the failure, and the CO₂ concentration could be estimated according to theenvironment at an absolute average error of 15.4 ppm. After that, it was applied to smart farmequipment to secure more reliable data and to maintain stable smart farm equipment, resulting in a lotof expected effects |
|
| 첨부논문 |
|
|
|
|
|
자료실 > 논문지 
