基于三向Dexel的GPU加速数控仿真方法  

A GPU ACCELERATED NC SIMULATION METHOD BASED ON TRI-DEXEL

作　　者：李一鹏 贾世宇 潘振宽 王吉强 李冰 Li Yipeng;Jia Shiyu;Pan Zhenkuan;Wang Jiqiang;Li Bing(College of Computer Science and Technology,Qingdao University,Qingdao 266000,Shandong,China)

机构地区：[1]青岛大学计算机科学技术学院,山东青岛266000

出　　处：《计算机应用与软件》2020年第11期63-68,共6页Computer Applications and Software

基　　金：国家十二五科技支撑计划项目(2013BAI01B03)

摘　　要：为了提升数控仿真的精度和速度,提出并行加速的基于三向Dexel的数控仿真方法。采用三向Dexel模型来表示工件与刀具,获得比使用单向Dexel模型的方法更真实的仿真效果。在表面重构方面,使用时间开销较大但效果更好的双轮廓法,通过使用并行化方法来解决仿真的实时性问题。仿真使用多线程加速和GPU加速技术,前者用于加速仿真中的布尔运算和三向Dexel数据转换过程,后者用于提高表面重构的速度。仿真实验结果表明,与仅适用CPU多线程的实现相比,应用GPU加速后表面重构的速度提高了5~15倍,CPU多线程的实现相比单线程的速度提高了232%~289%,能够满足数控仿真在精度和速度上的要求。To improve the accuracy and speed of NC-simulation,this paper presents a parallel accelerated NC simulation method based on Tri-dexel.It represented workpiece and tool instance with Tri-dexel models to obtain more authentic simulated results than traditional dexel.In the aspect of surface reconstruction,the double contouring,which costs more time but has better effect,was used,and the parallel method was used to solve the real-time problem of simulation.Multi-threading acceleration and GPU acceleration were used in the simulation.The former was used to accelerate Boolean operation and Tri-dexel data conversion process,while the latter was used to improve the speed of surface reconstruction.The simulation results show that compared with the implementation of CPU multi-threading only,the speed of surface reconstruction accelerated by GPU is 5~15 times,and the speed of CPU multi-threading is 232%~289%higher than that of single thread,which can meet the requirements of accuracy and speed of NC simulation.

关 键 词：数控加工仿真 三向深度元素 双轮廓 多线程加速 GPU加速 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]