Empirical study of CPS to support regional agriculture: Establishment of method to predict yield with high accuracy

2022 Fiscal Year Final Research Report

• Project/Area Number: 19K11986
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 60080:Database-related
• Research Institution: National Institute of Technology(KOSEN),Numazu College
• Principal Investigator: Yamazaki Satoshi 沼津工業高等専門学校, 制御情報工学科, 准教授 (80635889)
• Co-Investigator(Kenkyū-buntansha): 切岩 祥和 静岡大学, 農学部, 教授 (50303540)
• Project Period (FY): 2019-04-01 – 2023-03-31
• Keywords: 農業CPS / IoT / 省電力広域ネットワーク(LPWAN) / 施設栽培園芸作物 / イチゴ / 短期変動モデル / ノンパラメトリック回帰 / 収穫量予測

Outline of Final Research Achievements

There are two and main results of this study.
First, we proposed a crop yield prediction scheme using a cultivation model based on short-term fluctuations and generalized additive model for strawberries, which are horticultural crops that are harvested multiple times in one season. The proposed scheme showed higher prediction accuracy than the conventional method using actual IoT data. Second, we proposed several communication network methods to improve the energy efficiency of nodes focused on lower energy consumption and higher throughput when deploying IoT in a real environment. The effectiveness of the proposed methods were shown using either theoretical analysis, computer simulations, or hardware experiments.

Free Research Field

情報通信工学

Academic Significance and Societal Importance of the Research Achievements

本研究の学術的意義は、提案した栽培モデルとノンパラメトリック回帰を併用して施設栽培園芸作物の収穫量を予測する手法を提案し、（これまでほとんど栽培モデルが検討されてこなかった）イチゴに適用した際の有効性を定量的に示したことである。
本研究の社会的意義は、構築したIoTネットワークの実環境における性能を明らかにし、取得した実データを用いて従来手法と比べて提案手法が高い収穫量予測精度を示したことである。