金属增材制造负载均衡异构并行切片算法

Load Balancing Heterogeneous Parallel Slice Algorithm for Metal Additive Manufacturing

增材制造模型的几何复杂程度和体积不断提高,模型切片所需时间大幅增加,极大影响了数据处理效率。提出了一种自适应负载均衡的异构并行切片算法,利用不断增长的GPU超强并行计算能力对传统切片算法进行了GPU并行化研究,利用模拟退火算法对切片任务进行负载均衡装箱,使各线程间的任务量一致。实验结果证明,该算法提高了模型的切片效率,尤其适合大型或超大型三维模型的快速切片任务。

The geometric complexity and volume of additive manufacturing models continued to increase,and the time required for model slicing increased greatly,which greatly affected the efficiency of data processing.An adaptive load balancing heterogeneous parallel slicing algorithm was proposed.Using the growing parallel computing power of GPU,the traditional slicing algorithm was GPU parallelized,and the simulated annealing algorithm was used to load balance the slicing tasks,so that the task amount among the threads was consistent.Experimental results show that the algorithm improves the slicing efficiency of the model,and is especially suitable for fast slicing tasks of large or super large 3D models.