Abstract:

Objective: This study aims to address the issues of excessive reliance on complex and time-consuming outdoor experiments in the design and improvement of Camellia oleifera vibration harvesters, as well as the low simulation accuracy of traditional simulation models due to ignoring the mass of fruit-bud-leaf. A distributed mass based method for constructing a vibration model of C. oleifera trees was proposed-bud-leaf under vibration excitation, and to provide a theoretical basis for optimizing the harvester operational parameters. Method: First, statistical analysis was used to determine the distribution patterns of the number and mass of fruit-bud-leaf on the fruiting branches, and their total mass was applied as distributed mass to the fruit-bearing branches, thereby constructing a vibration model with fruit-bud-leaf. Next, hammer impact tests were conducted to obtain the natural frequencies of the trees, and vibration response tests were carried out to measure the acceleration at different positions of the trees. The test results were compared with simulation results under corresponding conditions to validate the accuracy of the model. Finally, based on the validated model, harmonic response analysis and field picking tests were conducted. By simulating the harmonic excitation response of trees under different excitation positions, the difficulty of fruit detachment was analyzed; through spectrum analysis, the optimal excitation frequency range for the harvesting device was determined. The best picking method was verified through field picking tests. Result: 1) Modal analysis and impact hammer test results showed that the first 15 natural frequencies simulated for by the C. oleifera tree model with fruit-bud-leaf were basically consistent with the impact hammer test results, with an average error of 6.64%. Compared with the tree model without fruit-bud-leaf, the average simulation error was reduced by 78.94%. 2) Transient analysis and vibration response test results indicated that the simulated acceleration values at 16 measurement points were in good agreement with the test acceleration values. The average values of the error, correlation coefficient, and acceleration ratio between the two were 24.06%, 0.85, and 1.03, respectively. Additionally, the amplitude of acceleration increased gradually along the branches from the bottom to the top with frequency. 3) Harmonic response analysis and field picking test results indicated fruits were more easily detached when the excitation force was applied to lateral branches and the excitation frequency ranged from 11.83 to 13.97 Hz. Conclusion: This paper establishes a vibration model for C. oleifera trees based on the mass distribution patterns of fruit-bud-leaf, effectively addressing the issue of low simulation accuracy in traditional models caused by neglecting the mass of fruit-bud-leaf. This model significantly improves the simulation accuracy. Furthermore, through harmonic response analysis and field picking tests, the excitation method for lateral branch harvesting of C. oleifera and the optimal excitation frequency range are determined. The research findings not only provide important theoretical guidance for the design and parameter optimization of C. oleifera harvesters but also offer a new method for constructing dynamics models for other forest fruit.

Key words: Camellia oleifera tree, vibration harvesting, distributed mass, fruit-bud-leaf model, simulation analysis