计算机集成制造系统 ›› 2019, Vol. 25 ›› Issue (第2): 303-317.DOI: 10.13196/j.cims.2019.02.004

• 当期目次 • 上一篇    下一篇

分布式环境下基于混合蛙跳算法的物化视图选择问题

陈于思,孙林夫+   

  1. 西南交通大学信息科学与技术学院
  • 出版日期:2019-02-28 发布日期:2019-02-28
  • 基金资助:
    国家重点研发计划资助项目 (2017YFB1400300)。

Shuffled frog leaping algorithm for view selection problem in distributed context

  • Online:2019-02-28 Published:2019-02-28
  • Supported by:
    Project supported by the National Key Research and Development Program,China (No.2017YFB1400300).

摘要: 为提高分布式环境下数据仓库的查询效率,降低维护成本,提出基于混合蛙跳算法(SFLA)求解分布式物化视图选择问题。分析了基本蛙跳规则不适用于分布式物化视图选择问题的原因,提出在局部搜索过程中,使用遗传算法重组算子替换基本蛙跳规则。扩展了遗传算法变异算子,以提高约束条件下的搜索能力,同时保持蛙群的多样性。提出启发式修复策略来处理进化过程中产生的不可行解。实验结果表明,在不同约束组合下,改进的SFLA在求解质量上优于基本SFLA和改进遗传算法;在约束较为严格时,从求解质量和稳定性的角度来看,启发式修复策略均明显优于惩罚策略和随机排名策略。

关键词: 物化视图, 分布式视图选择, 混合蛙跳算法, 进化算法

Abstract: To solve the View Selection Problem (VSP) under distributed scenario,a Shuffled Frog Leaping Algorithm (SFLA) based solution was presented,which was used to accelerate data warehouse queries through pre-computation and reduce the maintenance cost.After formalizing the DVSP model,the reason that the basic frog leaping rule was not applicable to DVSP was analyzed,and the method that using the recombination operator of genetic algorithm (GA) as an alternative was proposed in the local search procedure.The GA mutation operator was extended to improve the search ability under resource constraints while maintaining the diversity of frog population.A heuristic repair strategy was also proposed to deal with the infeasible solutions generated during the evolution.Experimental results showed the advantages of improved SFLA over basic SFLA and GA in terms of optimality under different constraints combinations.Compared with the penalty strategy and random ranking strategy,the heuristic repair strategy was superior in terms of solution quality and stability when the constraints were strict.

Key words: materialized view, distributed view selection, shuffled frog leaping algorithm, evolutionary algorithm

中图分类号: