Optimization of uprooting efficiency of counter-rotating cotton stalk puller for on-field operations

Background Cotton stalks, a by-product left after cotton picking, have several industrial applications as a raw material. However, due to deep taproot system, the uprooting and disposal of cotton stalks from the field is a labour-intensive operation. In this study, the uprooting efficiency of a counter-rotating drum type cotton stalk puller (CSP) was optimized using Response Surface Methodology (RSM) and combined Artificial Neural Network (ANN) - Particle Swarm Optimization (PSO) approach. Machine operational parameters and design parameter were independent variables, whereas, uprooting efficiency, plants broken and plants left were response variables. Results An experimental CSP unit was operated in field at three forward speeds (1.37, 1.67 and 1.95 km/h), four drum speeds (250, 300, 350 and 400 rpm) and three drum inclinations (0, 10, 20). The optimization using RSM shown 332.5 rpm drum speed, 8.36 drum inclination and 1.37 km/h forward speed as optimal values. Plants uprooted, plants broken and plants left have optimum values of 96.6%, 2.8% and 1.1% with individual desirability of 0.97, 0.85 and 0.89 showing the closeness of responses to predicted values. ANN-PSO model shown optimal parameters as 1.37 km/h forward speed, 7.89 drum inclination and 331.45 rpm drum speed with the observed and predicted values of uprooting efficiency are 96.72% and 94.84%, respectively. Conclusion The results show that both RSM and combined ANN-PSO approach can better predict and optimize the performance of CSP with higher accuracy. Optimization study provide essential information on optimal combination of operating and design parameters for enhanced uprooting efficiency with minimum plant breakage.