Efficient Packing of 2D Irregular Parts: A Hybrid Approach Incorporating a Modified Genetic Algorithm and Image Processing

This study proposes a technique for effectively arranging irregularly shaped parts on a board using a modified genetic algorithm and image processing. This technique addresses the challenge of efficiently packing parts of the same shape and size to optimize the utilization of available space. The optimization process comprises three search steps focused on finding suitable spatial relationships between the parts. The first two steps employ variance and envelope area criteria to optimize the position of the patterns, while the third step considers the distance between two columns of arranged parts. To enhance the accuracy and efficiency of the search process, a local-search-based optimization is proposed. The resulting optimized spatial relationships are derived from the three-step search process. The final layout strategy selects spatial relationships to maximize pattern accommodation on the board and arranges them horizontally to optimize utilization. The experimental results demonstrate the effectiveness of the proposed method in optimizing part layout for industrial production. Overall, this technique offers a solution for achieving effective packing, efficient resource utilization, and waste reduction.