Vegetable harvest assessment by analysis of vibrations
Date
2017Author
Jurkonis, Eugenijus
Stonkus, Rimantas
Dzedzickis, Andrius
Metadata
Show full item recordAbstract
Essential qualitative and quantitative harvest assessment is needed in order to plan and execute timely harvest when growing vining vegetables [1,2]. This is especially true in the harvesting by removing a few or several times during the fruiting season [3]. Usually, evaluation is carried out visually, and this takes time, but is not always successful because of lush foliage (especially in the cultivation of vegetables in the field) [4]. The paper proposes an assessment of harvest quantitative parameters using analysis of mechanical vibrations. Demonstration of the method is shown for the situation where support system in the field helps cucumbers grow vertically. Designed support construction digital model and calculated its natural vibrational frequencies by using FE method are proposed. In addition, analysis was done by modeling the foliage and / or cucumber yield. Most attention was paid to natural vibrational frequencies. The results were verified by laboratory experiment, actually simulating digital model. Measurements were done by taking advantage of the simple vibration recorders to capture oscillations of such a system as induced reaction to shock excitation. Counted and measured oscillation frequencies were recorded in the three test groups: - support model of cucumber vertically growing (numerical and experimental); - growing cucumber vertical support model and lush foliage; - cucumbers growing vertical support model, foliage and various harvest cases (corresponding to poor, medium and rich harvests). The method showed fairly significant changes in the lowest natural frequencies in different situations in the digital modeling and in experimental observation. Summarizing the results, it is possible to propose a use of such method for approximate harvest assessment and to manage best suitable times to collect cucumbers. The method can be adjusted using „machine learning\" elements for more accurate interpretation of the values of lowest natural vibration frequency variations.