Study On The Grasp Damage Of Harvesting Robot Based On The Biomechanical Properties Of Tomato Fruits

As the harvesting robot works, the robot fingers need to stable grasp a tomato fruit to seperate it from the stem. However, when the grasp forces that the fingers applied to tomato fruit exceed a threshold for tissue failure, they will cause severe mechanical damage to the fruit.when the forces are less than the minmum stable grasp force, the tomato fruit will slip out of the robot fingers and the drop damage will be caused. Different degrees of mechanical damage will lead to the fresh tomato fruits shrivel up, low quality and quickly rot. Therefore, it is essential for us to effectively prevent or reduce the tomato mechanical damage during mechanical harvest. In order to solve this problem, the stable grasp strategy of harvesting robot and these factors that affect fruit's grasp damage based on the biomechanical properties of tomato fruits were further investigated in this paper. The main research works are as follows:1) The biological properties of tomato fruits as related to robot's harvesting were investigated. It mainly includes the determination of tomato physical properties as related to mechanical harvesting, such as friction and rolling resistance coefficients, geometrical and morphological parameters, density and porosity; the ultrastructural observation of tomato skin, pericarp and locular gel tissues; the study of the effect of loading position and compressibility on mass loss and water content of damaged tomato. Results showed the loading position and compressibility have no significant effect on the water content of damaged tomato.loading position and compressibility have a gradual significant effect on mass loss of damaged tomato during storage at 12% and 16%.2) The mechanical properties of tomato fruits were investigated. It mainly includes the determination of mechanical properties of whole three and four locular tomato fruit, such as Ep, Fmax, rc and rk, and rupture mechanical properties such as Er, Fr andεby loading-unloading test; the determination of mechanical properties of tomato skin, pericarp and locular gel tissues by compression, tension, shear and bend test. Results showed the loading position has no significant effect on the whole mechanical properties of three locular tomato but has a significant effect on the partial whole mechanical properties of four locular tomato. The compressibility has a significant effect on the partial whole mechanical properties of four locular tomato.3) The finite element contact models were developed for harvesting robot fingers and tomato fruit. The plate fingers were replaced by rigid loading plates, and the whole tomato compression in CW and L position were simulated. In order to validate the developed models, the simulated results were compared with the experiment results in 3th chapter. Results showed the average relative errors of "average" model(FEA(Eavg)) are minimum under four loading conditions. Therefore, the finite element analysis model is proper for the prediction of grasp mechanical damage of tomato fruit when the elastic modulus of tomato skin, pericarp and locular gel tissues are average values respectively in the ANSYS large deformation static analysis.4) The effect of loading force and grasp position on the internal mechanical damage of tomato fruit were studied by nonlinear finite element analysis. Results showed the tomato locular gel, pericarp and skin tissues are damaged in order as the increasing of loading force under three loading methods. Under the same loading force, the compressibility is minimum when the four locular tomato is loaded by plate at CW, but the compressibility is maximum at L. The loading force and grasp position have a significant effect on the mechanical damage of tomato fruit. Meanwhile, The effect of fingers type on the interal mechanical damage of tomato fruit was also investigated.5) The stable grasp regions of harvesting robot finger on the equatorial region of tomato was predicted according to stable grasp theory. Results showed all the correct rate for prediction are more than 83.33%, meanwhile the reason of incorrect prediction was analysed. The logistic regression model of grasp mechanical damage is established using the data of loading-unloading test, and the factors that effect on the grasp mechanical damage of tomato fruit are discussed during robot harvesting. Results showed only compressibility, loading position(grasp position) and probe type(finger type) have a significant effect on the mechanical damage of tomato fruit for all the consideration factors, especially the compressibility.As a consequence, during robot's harvesting, firstly the geometrical properties of maximum vertical contour and equatorial contour are obtained by the vision system of harvesting robot. Subsequently,1) evaluate the tomato mass and calculate the grasp force applied the tomato fruit; 2) predict the stable grasp regions of harvesting robot finger on the equatorial region of tomato; 3) recognize the CW and L loading position of tomato fruit by the vision system. At last, let the robot fingers grasp the tomato from the L position. This grasp strategy is helpful to reduce the tomato mechanical damage during mechanical harvest.