Nondestructive and Fast Vibration Phenotyping of Plants

The frequencies of free oscillations of plants, or plant parts, depend on their geometries, stiffnesses, and masses. Besides direct biomechanical interest, free frequencies also provide insights into plant properties that can usually only be measured destructively or with low-throughput techniques (e.g., change in mass, tissue density, or stiffness over development or with stresses). We propose here a new high-throughput method based on the noncontact measurements of the free frequencies of the standing plant. The plant is excited by short air pulses (typically 100 ms). The resulting motion is recorded by a high speed video camera (100 fps) and processed using fast space and time correlation algorithms. In less than a minute the mechanical behavior of the plant is tested over several directions. The performance and versatility of this method has been tested in three contrasted species: tobacco (Nicotiana benthamian), wheat (Triticum aestivum L.), and poplar (Populus sp.), for a total of more than 4000 data points. In tobacco we show that water stress decreased the free frequency by 15%. In wheat we could detect variations of less than 1 g in the mass of spikes. In poplar we could measure frequencies of both the whole stem and leaves. The work provides insight into new potential directions for development of phenotyping.

The Robust NN Control and Fast Vibration Suppression of Space Flexible Manipulator with Unknown Parameters

To solve the problem of space flexible manipulator to track the desired trajectory, the robust control is proposed using a NN function to approach the comprehensive modeling errors. The adaptive feedback control for vibration suppression is designed of the system with unknown parameters. The adaptive laws are obtained by Lyapunov method without any prior knowledge about the bounds of unknown parameters. Finally, the numerical simulation is carried out, which proves the controller proposed is feasible and effective.