| Summary: | A harvesting mechanism for cabbage harvesters suitable in small-scale fields was developed in the laboratory. In the experiment, cutting speed, forward speed and cutting position were considered as independent parameters for the torque requirement to cut the cabbage stem in the harvesting process. The experiment was designed by full factorial design and ANN-PSO (Artificial Neural Network-Particle Swarm Optimization) technique was followed to optimize these cutting parameters to achieve minimum torque requirement for cutting. The cutting torque was found to be increased with an increase in the cutting position and a decrease in the cutting speed. An RBFNN (Radial Basis Function Neural Network) model was trained with 70% of the total cutting torque data, and the rest 30% of the data was used for testing the model. The mean square error (MSE) and coefficient of determination (R2) were obtained as 0.003 and 0.986, respectively which indicates the satisfactory performance of the model. The developed RBFNN-PSO model was validated against experimented data using four different combinations of input cutting parameters obtained from optimization. The optimized values of cutting speed (rpm), forward speed and cutting position were obtained as 590.75 rpm, 0.24 m s−1 and 0.825 cm, respectively with a deviation of 5.13%.
|
|---|