• CN:11-2187/TH
  • ISSN:0577-6686

›› 2011, Vol. 47 ›› Issue (13): 140-146.

• 论文 • 上一篇    下一篇

基于Kriging模型的复杂产品管线敷设顺序粒子群优化

柳强;王成恩   

  1. 东北大学辽宁省复杂装备多学科设计优化技术重点实验室;东北大学流程工业综合自动化教育部重点实验室
  • 发布日期:2011-07-05

Kriging Model-based Routing Sequence Planning for Complex Products by Particle Swarm Optimization

LIU Qiang; WANG Chengen   

  1. Liaoning Province Key Laboratory of Multidisciplinary Optimal Design for Complex Equipment, Northeastern University Key Laboratory of Integrated Automation of Process Industry of Ministry of Education, Northeastern University
  • Published:2011-07-05

摘要: 复杂产品的管线布局问题十分复杂,除管路路径规划算法外,敷设顺序对管线整体敷设有着十分重要的影响。以往的研究大多将重点集中在管线路径规划方面,而对敷设顺序的研究还比较少。以航空发动机为例,提出一种基于Kriging模型的管线敷设顺序的规划方法。针对敷设顺序规划这一离散优化问题,通过论域连续化处理,建立粒子群算法优化相关函数参数的Kriging近似模型。在Kriging模型的基础上,应用离散粒子群算法搜索最优顺序,避免在优化过程中反复应用管线路径规划算法进行计算,显著提高规划效率。对离散粒子群算法的改进进一步提升算法的搜索性能。所提规划方法具有很好的通用性,仿真试验验证了方法的有效性与高效性。

关键词: Kriging模型, 敷设顺序, 管路敷设, 粒子群优化

Abstract: Pipe routing for complex products is very complicated. Besides the pipe path planning algorithm, the routing sequence also exerts remarkable impacts on final pipe layouts. While most previous researchers focus on seeking the pipe paths, few efforts are made on routing sequence planning. The routing sequence planning problem is studied in the context of an aero-engine development, and a Kriging model-based routing sequence planning algorithm is presented. In order to handle the discrete planning problem, a discrete Kriging model with correlation functions optimized by particle swarm optimization is constructed by using a continuous handling method. Based on the constructed Kriging model, the optimal pipe routing sequence is determined by using a discrete particle swarm optimization, which dramatically improves computation efficiency by replacing the path planning algorithm with the Kriging model. The discrete particle swarm optimization is modified to improve the search ability. The proposed planning algorithm has good universality, and the effectiveness and efficiency of which are demonstrated by the simulations.

Key words: Kriging model, Particle swarm optimization, Pipe routing, Routing sequences

中图分类号: