Adaptive Genetic Algorithm for Acid Rolling Scheduling Problem in the Steel Industry

For an acid rolling scheduling problem arising from steel production system, a mathematical programming model was established to transform the violation states or jump values of adjacent materials' width, thickness, steel type, welding codes, etc., during the production process into transition costs between materials. An adaptive genetic algorithm (AGA) was proposed to address the characteristics of the problem. It utilizes an adaptive parameter control method to adjust penalty values for attributes with a higher number of violations, thus ensuring the algorithm’s applicability to a wide range of units. The model and algorithm were validated using a practical production case in a specific enterprise. Experimental results demonstrated that AGA reduces the violations in scheduling plans while lowering the jumps in attributes such as the width and thickness of adjacent materials. Consequently, it achieves the objectives of reducing unit switching times, decreasing production costs, and improving production efficiency.