Hybrid ANN-GA Approach for Maximizing the Capacity of a Screw Conveyor: Experimental Investigation and Parametric Optimization

This study presents a comprehensive optimization of screw conveyor performance using artificial intelligence techniques. A laboratory-scale screw conveyor system was designed with the following specifications: diameter (D) = ۰.۲ m, pitch (s) = ۰.۱۵ m, length = ۲ m, constructed from stainless steel, driven by a variable-speed motor (۶۰۰-۱۲۰۰ rpm) with digital control, and featuring an adjustable tilt frame (۰-۸۰°) to cover industrial applications. Bulk wheat flour with density p = ۶۰۰ kg/m³ was used as the transported material. Experimental investigations were conducted to examine the effects of rotational speed (rpm) and inclination angle (°) on conveyor capacity (ton/h). For each test, ۴۰ kg of wheat was transported, with transfer time recorded using a stopwatch. Capacity values were calculated and recorded in ton/h. The collected dataset was analyzed using both Artificial Neural Networks (ANN) and Genetic Algorithm (GA) approaches. The ANN model demonstrated superior predictive capability with R۲ = ۰.۸۹۴ and RMSE = ۰.۲۶۷, effectively capturing the complex nonlinear relationships between operational parameters and conveyor capacity. Subsequent optimization using GA identified the optimal operating conditions: inclination angle of ۱۹.۸° and rotational speed of ۱۰۷۲ rpm, yielding a predicted maximum capacity of ۴.۸۷ ton/h. This represents an ۸۴.۵% improvement over average performance conditions. Comparative analysis revealed that the hybrid ANN-GA approach outperformed traditional polynomial regression methods (R۲ = ۰.۸۳۲), providing more accurate predictions and reliable optimization results. The findings demonstrate that intelligent optimization techniques can significantly enhance screw conveyor efficiency, with practical implications for industrial material handling systems where optimal parameter selection is crucial for energy efficiency and operational performance.

کلیدواژه ها:

Screw Conveyor ، Optimization ، Artificial Neural Networks ، Genetic Algorithm ، Bulk Material Handling ، Capacity Optimization ، Industrial Automation