The Tempo machine plants seeds at controlled intervals using the PowerShoot system. Model-based diagnosis utilizes relationships between measured states and system model estimations to detect and isolate faults. This thesis, conducted in collaboration with Väderstad AB, focuses on developing a model-based diagnosis system for the Tempo L. A system model was created and validated using data from a dedicated testing machine, demonstrating the ability to estimate most machine states from hydraulic pressure as the sole input. Using the validated models, the MATLAB Fault Diagnosis Toolbox was employed to generate state comparisons and residuals linked to common faults. Residuals were evaluated using CUSUM tests to detect faults while minimizing false alarms. The diagnosis system identified 15 faults, of which 13 were detectable and 9 fully isolable, with major faults including leakages and blockages affecting seeding quality. The thesis also explores the use of optical sensors in each row unit to predict pressure based on seed velocity, presenting promising diagnostic abilities. Additionally, different sensor configurations were evaluated to assess the benefits of adding more sensors to the machine. This work represents an initial effort to improve the diagnostics of the Tempo L planter, providing insight into more reliable and efficient agricultural machinery.
Identifer | oai:union.ndltd.org:UPSALLA1/oai:DiVA.org:liu-204989 |
Date | January 2024 |
Creators | Ederlöv, Isak, Mineur, Viktor |
Publisher | Linköpings universitet, Fordonssystem |
Source Sets | DiVA Archive at Upsalla University |
Language | English |
Detected Language | English |
Type | Student thesis, info:eu-repo/semantics/bachelorThesis, text |
Format | application/pdf |
Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0018 seconds