Structural Fault Detection and Diagnosis for Combine Harvesters: A Critical Review

Area of Science:

• Agricultural Engineering
• Machine Learning
• Data Science

Background:

• Combine harvesters face frequent structural faults due to complex designs and harsh operating conditions.
• These faults reduce operational efficiency, cause downtime, and impact agricultural productivity, threatening food security.

Purpose of the Study:

• To systematically review and analyze recent advancements in data-driven Fault Detection and Diagnosis (FDD) methods for structural faults in combine harvesters.
• To provide insights into implementing these FDD methods and identify current challenges and future research directions.

Main Methods:

• Review of data-driven FDD techniques, including sensor data acquisition (vibration, acoustic, strain), signal preprocessing, and feature extraction (time, frequency, time-frequency domains, modal analysis).
• Analysis of machine learning and artificial intelligence models for fault pattern recognition and diagnosis.
• Exploration of system and technical support, such as on-board diagnostics, remote monitoring, and simulation modeling.

Main Results:

• Identification of common structural faults in combine harvesters and their typical locations.
• Detailed overview of the data-driven FDD process from data acquisition to diagnosis.
• Discussion on system requirements and technical support for practical FDD implementation.

Conclusions:

• Data-driven FDD methods show significant promise for enhancing combine harvester reliability and operational efficiency.
• Key challenges include data acquisition difficulties, signal complexity, and model robustness, necessitating further research.
• Future directions aim to develop intelligent maintenance strategies for agricultural machinery.