Development and performance evaluation of variable width raised bed former with optimal parameters predicted by RBF...

Area of Science:

• Agricultural Engineering
• Soil Science
• Computational Modeling

Background:

• Raised bed planting enhances soil health, water efficiency, and energy use.
• Crop-specific needs require adaptable machinery like variable width bed formers.

Purpose of the Study:

• To develop and evaluate a variable width raised bed former.
• To optimize its operational parameters using advanced computational methods.

Main Methods:

• Soil bin experiments analyzed draft force under varying soil moisture, working width, and forward speed.
• Regression and Radial Basis Function (RBF) neural networks modeled specific draft.
• RBF neural network-Particle Swarm Optimization (PSO) technique optimized input parameters.

Main Results:

• Both regression and RBF models accurately predicted specific draft (R² > 0.98).
• Optimal parameters yielded specific draft close to actual values in soil bin tests.
• Field tests showed a ±7.2% variation between predicted and actual specific draft.

Conclusions:

• Variable width raised bed formers are effective for diverse crop needs.
• Neural network and PSO techniques successfully optimize tillage machinery parameters.
• These methods reduce energy expenditure in agricultural field operations.